Variability and Standardization of Quantitative Imaging: Monoparametric to Multiparametric Quantification, Radiomics, and Artificial Intelligence

Repeatability of 3D MR fingerprinting during scanner software upgrades

OBJECTIVE

This study aims to quantify the repeatability of a 3D Magnetic Resonance Fingerprinting (MRF) research protocol in the context of a scanner software upgrade. All of MRI assumes consistent hardware performance and raw data pre-processing on the acquisition side. Software upgrades can affect hardware specifications and reconstruction chain parameters. Understanding how vendor-provided software upgrades vary MRF-derived T1 and T2 values is crucial for its application in different settings.

MATERIALS AND METHODS

Eight healthy volunteers were imaged with an in-house developed 3D MRF pulse sequence using a 3T scanner before and after a software upgrade (VA31A to VA50A, MAGNETOM Vida, Siemens Healthineers). Online MRF reconstruction using Singular Value Decomposition (SVD) timeseries compression and B1+ correction was performed. The study involved test-retest repeatability assessment and a comparison of pre- and post-upgrade data based on automatically extracted T1 and T2 values from MNI-152 Harvard-Oxford Subcortical Structural Atlas regions.

RESULTS

Significant mismatches were found directly after the upgrade. However, after an information exchange with the vendor, the 3D-MRF sequence showed consistent repeatability in both intra-version test-retest scenarios and cross-version comparisons: - 1.16 ± 3.18% variability in T1 and - 0.54 ± 4.84% in T2 for intra-version tests, and - 0.83 ± 3.68% (T1) and - 0.05 ± 5.81% (T2) variability for cross-version comparisons.

DISCUSSION

The study shows the reliable performance of 3D MRF protocols across software upgrades is possible, but it also highlights the importance of detailed evaluation and vendor collaboration in ensuring consistency. These findings support the application of MRF in longitudinal studies and emphasize the need for systematic assessments following hardware or software modifications.

Three-dimensional simultaneous T1 and T2* relaxation times and quantitative susceptibility mapping at 3 T: A multicenter validation study

We aimed to determine the intra-site repeatability and cross-site reproducibility of T1 and T2* relaxation times and quantitative susceptibility (χ) values obtained through quantitative parameter mapping (QPM) at 3 T. This prospective study included three 3-T scanners with the same hardware and software platform at three sites. The brains of twelve healthy volunteers were scanned three times using QPM at three sites. Intra-site repeatability and cross-site reproducibility were evaluated based on voxel-wise and region-of-interest analyses. The within-subject coefficient of variation (wCV), within-subject standard deviation (wSD), linear regression, Bland-Altman plot, and intraclass correlation coefficient (ICC) were used for evaluation. The intra-site repeatability wCV was 11.9 ± 6.86% for T1 and 3.15 ± 0.03% for T2*, and wSD of χ at 3.35 ± 0.10 parts per billion (ppb). Intra-site ICC(1,k) values for T1, T2*, and χ were 0.878-0.904, 0.972-0.976, and 0.966-0.972, respectively, indicating high consistency within the same scanner. Linear regression analysis revealed a strong agreement between measurements from each site and the site-average measurement, with R-squared values ranging from 0.79 to 0.83 for T1, 0.94-0.95 for T2*, and 0.95-0.96 for χ. The cross-site wCV was 13.4 ± 5.47% for T1 and 3.69 ± 2.25% for T2*, and cross-site wSD of χ at 4.08 ± 3.22 ppb. The cross-site ICC(2,1) was 0.707, 0.913, and 0.902 for T1, T2*, and χ, respectively. QPM provides T1, T2*, and χ values with an intra-site repeatability of <12% and cross-site reproducibility of <14%. These findings may contribute to the development of multisite studies.

radMLBench: A dataset collection for benchmarking in radiomics

BACKGROUND

New machine learning methods and techniques are frequently introduced in radiomics, but they are often tested on a single dataset, which makes it challenging to assess their true benefit. Currently, there is a lack of a larger, publicly accessible dataset collection on which such assessments could be performed. In this study, a collection of radiomics datasets with binary outcomes in tabular form was curated to allow benchmarking of machine learning methods and techniques.

METHODS

A variety of journals and online sources were searched to identify tabular radiomics data with binary outcomes, which were then compiled into a homogeneous data collection that is easily accessible via Python. To illustrate the utility of the dataset collection, it was applied to investigate whether feature decorrelation prior to feature selection could improve predictive performance in a radiomics pipeline.

RESULTS

A total of 50 radiomic datasets were collected, with sample sizes ranging from 51 to 969 and 101 to 11165 features. Using this data, it was observed that decorrelating features did not yield any significant improvement on average.

CONCLUSIONS

A large collection of datasets, easily accessible via Python, suitable for benchmarking and evaluating new machine learning techniques and methods was curated. Its utility was exemplified by demonstrating that feature decorrelation prior to feature selection does not, on average, lead to significant performance gains and could be omitted, thereby increasing the robustness and reliability of the radiomics pipeline.

Linac- and CyberKnife-based MRI-only treatment planning of prostate SBRT using an optimized synthetic CT calibration curve

PURPOSE

CT Hounsfield Units (HUs) are converted to electron density using a calibration curve obtained from physical measurements of an electron density phantom. HU values assigned to an MRI-derived synthetic computed tomography (sCT) may present a different relationship with electron density compared to CT HU. Correct assignment of sCT HU values is critical for accurate dose calculation and delivery. The goals of this work were to develop a sCT calibration curve using patient data acquired on a clinically commissioned CT scanner and assess for CyberKnife- and volumetric modulated arc therapy (VMAT)-based MR-only treatment planning of prostate SBRT.

METHODS

Same-day CT and MRI simulation in the treatment position were performed on 10 patients treated with SBRT to the prostate. Dixon in-phase and out-of-phase MRIs were acquired on a 3T scanner using a 3D T1-weighted gradient-echo sequence to generate sCTs using a commercial sCT algorithm. CT and sCT datasets were co-registered and HU values compared using mean absolute error (MAE). An optimized HU-to-density calibration curve was created based on average HU values across an institutional patient database for each of the four sCT tissue types. Clinical CyberKnife and VMAT treatment plans were generated on each patient CT and recomputed onto corresponding sCTs. Dose distributions computed using CT and sCT were compared using gamma criteria and dose-volume-histograms.

RESULTS

For the optimized calibration curve, HU values were -96, 37, 204, and 1170 and relative electron densities were 0.95, 1.04, 1.1, and 1.7 for adipose, soft tissue, inner bone, and outer bone, respectively. The proposed sCT protocol produced total MAE of 94 ± 20HU. Gamma values mean ± std (min-max) were 98.9% ± 0.9% (97.1%-100%) and 97.7% ± 1.3% (95.3%-99.3%) for VMAT and CyberKnife plans, respectively.

CONCLUSION

MRI-derived sCT using the proposed approach shows excellent dosimetric agreement with conventional CT simulation, demonstrating the feasibility of MRI-derived sCT for prostate SBRT treatment planning.

Oncologic Applications of Artificial Intelligence and Deep Learning Methods in CT Spine Imaging-A Systematic Review

In spinal oncology, integrating deep learning with computed tomography (CT) imaging has shown promise in enhancing diagnostic accuracy, treatment planning, and patient outcomes. This systematic review synthesizes evidence on artificial intelligence (AI) applications in CT imaging for spinal tumors. A PRISMA-guided search identified 33 studies: 12 (36.4%) focused on detecting spinal malignancies, 11 (33.3%) on classification, 6 (18.2%) on prognostication, 3 (9.1%) on treatment planning, and 1 (3.0%) on both detection and classification. Of the classification studies, 7 (21.2%) used machine learning to distinguish between benign and malignant lesions, 3 (9.1%) evaluated tumor stage or grade, and 2 (6.1%) employed radiomics for biomarker classification. Prognostic studies included three (9.1%) that predicted complications such as pathological fractures and three (9.1%) that predicted treatment outcomes. AI's potential for improving workflow efficiency, aiding decision-making, and reducing complications is discussed, along with its limitations in generalizability, interpretability, and clinical integration. Future directions for AI in spinal oncology are also explored. In conclusion, while AI technologies in CT imaging are promising, further research is necessary to validate their clinical effectiveness and optimize their integration into routine practice.

Novel spin-lock time sampling strategies for improved reproducibility in quantitative T1ρ mapping

This study aimed to optimize the sampling of spin-lock times (TSLs) in quantitative T1ρ mapping for improved reproducibility. Two new TSL sampling schemes were proposed: (i) reproducibility-guided random sampling (RRS) and (ii) reproducibility-guided optimal sampling (ROS). They were compared to the existing linear sampling (LS) and precision-guided sampling (PS) schemes for T1ρ reproducibility through numerical simulations, phantom experiments, and volunteer studies. Each study evaluated the four sampling schemes with three commonly used T1ρ preparations based on composite and balanced spin-locking. Additionally, the phantom and volunteer studies investigated the impact of B0 and B1 field inhomogeneities on T1ρ reproducibility, respectively. The reproducibility was assessed using the coefficient of variation (CoV) by repeating the T1ρ measurements eight times for phantom experiments and five times for volunteer studies. Numerical simulations resulted in lower mean CoVs for the proposed RRS (1.74%) and ROS (0.68%) compared to LS (2.93%) and PS (3.68%). In the phantom study, the mean CoVs were also lower for RRS (2.7%) and ROS (2.6%) compared to LS (4.1%) and PS (3.1%). Furthermore, the mean CoVs of the proposed RRS and ROS were statistically lower (P < 0.001) compared to existing LS and PS schemes at a B1 offset of 20%. In the volunteer study, consistently lower mean CoVs were observed in bilateral thigh muscles for RRS (9.3%) and ROS (9.2%) compared to LS (10.9%) and PS (10.2%), and the difference was more prominent at B0 offsets higher than 50 Hz. The proposed sampling schemes improve the reproducibility of quantitative T1ρ mapping by optimizing the selection of TSLs. This improvement is especially beneficial for longitudinal studies that track and monitor disease progression and treatment response.

A Comprehensive Review on the Role of MRI in the Assessment of Supratentorial Neoplasms: Comparative Insights Into Adult and Pediatric Cases

Magnetic resonance imaging (MRI) is a critical diagnostic tool in assessing supratentorial neoplasms, offering unparalleled detail and specificity in brain imaging. Supratentorial neoplasms in the cerebral hemispheres, basal ganglia, thalamus, and other structures above the tentorium cerebelli present significant diagnostic and therapeutic challenges. These challenges vary notably between adult and pediatric populations due to differences in tumor types, biological behavior, and patient management strategies. This comprehensive review explores the role of MRI in diagnosing, planning treatment, monitoring response, and detecting recurrence in supratentorial neoplasms, providing comparative insights into adult and pediatric cases. The review begins with an overview of the epidemiology and pathophysiology of these tumors in different age groups, followed by a detailed examination of standard and advanced MRI techniques, including diffusion-weighted imaging (DWI), perfusion-weighted imaging (PWI), and magnetic resonance spectroscopy (MRS). We discuss the specific imaging characteristics of various neoplasms and the importance of tailored approaches to optimize diagnostic accuracy and therapeutic efficacy. The review also addresses the technical and interpretative challenges unique to pediatric imaging and the implications for long-term patient outcomes. By highlighting the comparative utility of MRI in adult and pediatric cases, this review aims to enhance the understanding of its pivotal role in managing supratentorial neoplasms. It underscores the necessity of age-specific diagnostic and therapeutic strategies. Emerging MRI technologies and future research directions are also discussed, emphasizing the potential for advancements in personalized imaging approaches and improved patient care across all age groups.

Advancing Medical Imaging Research Through Standardization: The Path to Rapid Development, Rigorous Validation, and Robust Reproducibility

ABSTRACT

Artificial intelligence (AI) has made significant advances in radiology. Nonetheless, challenges in AI development, validation, and reproducibility persist, primarily due to the lack of high-quality, large-scale, standardized data across the world. Addressing these challenges requires comprehensive standardization of medical imaging data and seamless integration with structured medical data.Developed by the Observational Health Data Sciences and Informatics community, the OMOP Common Data Model enables large-scale international collaborations with structured medical data. It ensures syntactic and semantic interoperability, while supporting the privacy-protected distribution of research across borders. The recently proposed Medical Imaging Common Data Model is designed to encompass all DICOM-formatted medical imaging data and integrate imaging-derived features with clinical data, ensuring their provenance.The harmonization of medical imaging data and its seamless integration with structured clinical data at a global scale will pave the way for advanced AI research in radiology. This standardization will enable federated learning, ensuring privacy-preserving collaboration across institutions and promoting equitable AI through the inclusion of diverse patient populations. Moreover, it will facilitate the development of foundation models trained on large-scale, multimodal datasets, serving as powerful starting points for specialized AI applications. Objective and transparent algorithm validation on a standardized data infrastructure will enhance reproducibility and interoperability of AI systems, driving innovation and reliability in clinical applications.

Automatic cephalometric landmark identification with artificial intelligence: An umbrella review of systematic reviews

OBJECTIVES

The transition from manual to automatic cephalometric landmark identification has not yet reached a consensus for clinical application in orthodontic diagnosis. The present umbrella review aimed to assess artificial intelligence (AI) performance in automatic 2D and 3D cephalometric landmark identification.

DATA

A combination of free text words and MeSH keywords pooled by boolean operators: Automa* AND cephalo* AND ("artificial intelligence" OR "machine learning" OR "deep learning" OR "learning").

SOURCES

A search strategy without a timeframe setting was conducted on PubMed, Scopus, Web of Science, Cochrane Library and LILACS.

STUDY SELECTION

The study protocol followed the PRISMA guidelines and the PICO question was formulated according to the aim of the article. The database search led to the selection of 15 articles that were assessed for eligibility in full-text. Finally, 11 systematic reviews met the inclusion criteria and were analyzed according to the risk of bias in systematic reviews (ROBIS) tool.

CONCLUSIONS

AI was not able to identify the various cephalometric landmarks with the same accuracy. Since most of the included studies' conclusions were based on a wrong 2 mm cut-off difference between the AI automatic landmark location and that allocated by human operators, future research should focus on refining the most powerful architectures to improve the clinical relevance of AI-driven automatic cephalometric analysis.

CLINICAL SIGNIFICANCE

Despite a progressively improved performance, AI has exceeded the recommended magnitude of error for most cephalometric landmarks. Moreover, AI automatic landmarking on 3D CBCT appeared to be less accurate compared to that on 2D X-rays. To date, AI-driven cephalometric landmarking still requires the final supervision of an experienced orthodontist.

Influence of image preprocessing on the segmentation-based reproducibility of radiomic features: in vivo experiments on discretization and resampling parameters

PURPOSE

To systematically investigate the impact of image preprocessing parameters on the segmentation-based reproducibility of magnetic resonance imaging (MRI) radiomic features.

METHODS

The MRI scans of 50 patients were included from the multi-institutional Brain Tumor Segmentation 2021 public glioma dataset. Whole tumor volumes were manually segmented by two independent readers, with the participation of eight readers. Radiomic features were extracted from two sequences: T2-weighted (T2) and contrast-enhanced T1-weighted (T1ce). Two methods were considered for discretization: bin count (i.e., relative discretization) and bin width (i.e., absolute discretization). Ten discretization (five for each method) and five resampling parameters were varied while other parameters were fixed. The intraclass correlation coefficient (ICC) was used for reliability analysis based on two commonly used cut-off values (0.75 and 0.90).

RESULTS

Image preprocessing parameters had a significant impact on the segmentation-based reproducibility of radiomic features. The bin width method yielded more reproducible features than the bin count method. In discretization experiments using the bin width on both sequences, according to the ICC cut-off values of 0.75 and 0.90, the rate of reproducible features ranged from 70% to 84% and from 35% to 57%, respectively, with an increasing percentage trend as parameter values decreased (from 84 to 5 for T2; 100 to 6 for T1ce). In the resampling experiments, these ranged from 53% to 74% and from 10% to 20%, respectively, with an increasing percentage trend from lower to higher parameter values (physical voxel size; from 1 x 1 x 1 to 2 x 2 x 2 mm3).

CONCLUSION

The segmentation-based reproducibility of radiomic features appears to be substantially influenced by discretization and resampling parameters. Our findings indicate that the bin width method should be used for discretization and lower bin width and higher resampling values should be used to allow more reproducible features.

Evaluating a Novel AI Tool for Automated Measurement of the Aortic Root and Valve in Cardiac Magnetic Resonance Imaging

Objective Evaluating an artificial intelligence (AI) tool (AIATELLA, version 1.0; AIATELLA Oy, Helsinki, Finland) in interpreting cardiac magnetic resonance (CMR) imaging to produce measurements of the aortic root and valve by comparison of accuracy and efficiency with that of three National Health Service (NHS) cardiologists. Methods AI-derived aortic root and valve measurements were recorded alongside manual measurements from three experienced NHS consultant cardiologists (CCs) over three separate sites in the northeast part of the United Kingdom. The study utilised a comprehensive dataset of CMR images, with the intraclass correlation coefficient (ICC) being the primary measure of concordance between the AI and the cardiologist assessments. Patient imaging was anonymised and blinded at the point of transfer to a secure data server.  Results The study demonstrates a high level of concordance between AI assessment of the aortic root and valve with NHS cardiologists (ICC of 0.98). Notably, the AI delivered results in 2.6 seconds (+/- 0.532) compared to a mean of 334.5 seconds (+/- 61.9) by the cardiologists, a statistically significant improvement in efficiency without compromising accuracy. Conclusion AI's accuracy and speed of analysis suggest that it could be a valuable tool in cardiac diagnostics, addressing the challenges of time-consuming and variable clinician-based assessments. This research reinforces AI's role in optimising the patient journey and improving the efficiency of the diagnostic pathway.

Cross-vendor multiparametric mapping of the human brain using 3D-QALAS: A multicenter and multivendor study

PURPOSE

To evaluate a vendor-agnostic multiparametric mapping scheme based on 3D quantification using an interleaved Look-Locker acquisition sequence with a T2 preparation pulse (3D-QALAS) for whole-brain T1, T2, and proton density (PD) mapping.

METHODS

This prospective, multi-institutional study was conducted between September 2021 and February 2022 using five different 3T systems from four prominent MRI vendors. The accuracy of this technique was evaluated using a standardized MRI system phantom. Intra-scanner repeatability and inter-vendor reproducibility of T1, T2, and PD values were evaluated in 10 healthy volunteers (6 men; mean age ± SD, 28.0 ± 5.6 y) who underwent scan-rescan sessions on each scanner (total scans = 100). To evaluate the feasibility of 3D-QALAS, nine patients with multiple sclerosis (nine women; mean age ± SD, 48.2 ± 11.5 y) underwent imaging examination on two 3T MRI systems from different manufacturers.

RESULTS

Quantitative maps obtained with 3D-QALAS showed high linearity (R2  = 0.998 and 0.998 for T1 and T2, respectively) with respect to reference measurements. The mean intra-scanner coefficients of variation for each scanner and structure ranged from 0.4% to 2.6%. The mean structure-wise test-retest repeatabilities were 1.6%, 1.1%, and 0.7% for T1, T2, and PD, respectively. Overall, high inter-vendor reproducibility was observed for all parameter maps and all structure measurements, including white matter lesions in patients with multiple sclerosis.

CONCLUSION

The vendor-agnostic multiparametric mapping technique 3D-QALAS provided reproducible measurements of T1, T2, and PD for human tissues within a typical physiological range using 3T scanners from four different MRI manufacturers.

Diffusion tensor imaging in anisotropic tissues: application of reduced gradient vector schemes in peripheral nerves

BACKGROUND

In contrast to the brain, fibers within peripheral nerves have distinct monodirectional structure questioning the necessity of complex multidirectional gradient vector schemes for DTI. This proof-of-concept study investigated the diagnostic utility of reduced gradient vector schemes in peripheral nerve DTI.

METHODS

Three-Tesla magnetic resonance neurography of the tibial nerve using 20-vector DTI (DTI20) was performed in 10 healthy volunteers, 12 patients with type 2 diabetes, and 12 age-matched healthy controls. From the full DTI20 dataset, three reduced datasets including only two or three vectors along the x- and/or y- and z-axes were built to calculate major parameters. The influence of nerve angulation and intraneural connective tissue was assessed. The area under the receiver operating characteristics curve (ROC-AUC) was used for analysis.

RESULTS

Simplified datasets achieved excellent diagnostic accuracy equal to DTI20 (ROC-AUC 0.847-0.868, p ≤ 0.005), but compared to DTI20, the reduced models yielded mostly lower absolute values of DTI scalars: median fractional anisotropy (FA) ≤ 0.12; apparent diffusion coefficient (ADC) ≤ 0.25; axial diffusivity ≤ 0.96, radial diffusivity ≤ 0.07). The precision of FA and ADC with the three-vector model was closest to DTI20. Intraneural connective tissue was negatively correlated with FA and ADC (r ≥ -0.49, p < 0.001). Small deviations of nerve angulation had little effect on FA accuracy.

CONCLUSIONS

In peripheral nerves, bulk tissue DTI metrics can be approximated with only three predefined gradient vectors along the scanner's main axes, yielding similar diagnostic accuracy as a 20-vector DTI, resulting in substantial scan time reduction.

RELEVANCE STATEMENT

DTI bulk tissue parameters of peripheral nerves can be calculated with only three predefined gradient vectors at similar diagnostic performance as a standard DTI but providing a substantial scan time reduction.

KEY POINTS

• In peripheral nerves, DTI parameters can be approximated using only three gradient vectors. • The simplified model achieves a similar diagnostic performance as a standard DTI. • The simplified model allows for a significant acceleration of image acquisition. • This can help to introduce multi-b-value DTI techniques into clinical practice.

Achieving imaging and computational reproducibility on multiparametric MRI radiomics features in brain tumor diagnosis: phantom and clinical validation

OBJECTIVES

The Image Biomarker Standardization Initiative has helped improve the computational reproducibility of MRI radiomics features. Nonetheless, the MRI sequences and features with high imaging reproducibility are yet to be established. To determine reproducible multiparametric MRI radiomics features across test-retest, multi-scanner, and computational reproducibility comparisons, and to evaluate their clinical value in brain tumor diagnosis.

METHODS

To assess reproducibility, T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), and diffusion-weighted imaging (DWI) were acquired from three 3-T MRI scanners using standardized phantom, and radiomics features were extracted using two computational algorithms. Reproducible radiomics features were selected when the concordance correlation coefficient value above 0.9 across multiple sessions, scanners, and computational algorithms. Random forest classifiers were trained with reproducible features (n = 117) and validated in a clinical cohort (n = 50) to evaluate whether features with high reproducibility improved the differentiation of glioblastoma from primary central nervous system lymphomas (PCNSLs).

RESULTS

Radiomics features from T2WI demonstrated higher repeatability (65-94%) than those from DWI (38-48%) or T1WI (2-92%). Across test-retest, multi-scanner, and computational comparisons, T2WI provided 41 reproducible features, DWI provided six, and T1WI provided two. The performance of the classification model with reproducible features was higher than that using non-reproducible features in both training set (AUC, 0.916 vs. 0.877) and validation set (AUC, 0.957 vs. 0.869).

CONCLUSION

Radiomics features with high reproducibility across multiple sessions, scanners, and computational algorithms were identified, and they showed higher diagnostic performance than non-reproducible radiomics features in the differentiation of glioblastoma from PCNSL.

CLINICAL RELEVANCE STATEMENT

By identifying the radiomics features showing higher multi-machine reproducibility, our results also demonstrated higher radiomics diagnostic performance in the differentiation of glioblastoma from PCNSL, paving the way for further research designs and clinical application in neuro-oncology.

KEY POINTS

• Highly reproducible radiomics features across multiple sessions, scanners, and computational algorithms were identified using phantom and applied to clinical diagnosis. • Radiomics features from T2-weighted imaging were more reproducible than those from T1-weighted and diffusion-weighted imaging. • Radiomics features with good reproducibility had better diagnostic performance for brain tumors than features with poor reproducibility.

Deep Learning-Based Kernel Adaptation Enhances Quantification of Emphysema on Low-Dose Chest CT for Predicting Long-Term Mortality

OBJECTIVES

The aim of this study was to ascertain the predictive value of quantifying emphysema using low-dose computed tomography (LDCT) post deep learning-based kernel adaptation on long-term mortality.

MATERIALS AND METHODS

This retrospective study investigated LDCTs obtained from asymptomatic individuals aged 60 years or older during health checkups between February 2009 and December 2016. These LDCTs were reconstructed using a 1- or 1.25-mm slice thickness alongside high-frequency kernels. A deep learning algorithm, capable of generating CT images that resemble standard-dose and low-frequency kernel images, was applied to these LDCTs. To quantify emphysema, the lung volume percentage with an attenuation value less than or equal to -950 Hounsfield units (LAA-950) was gauged before and after kernel adaptation. Low-dose chest CTs with LAA-950 exceeding 6% were deemed emphysema-positive according to the Fleischner Society statement. Survival data were sourced from the National Registry Database at the close of 2021. The risk of nonaccidental death, excluding causes such as injury or poisoning, was explored according to the emphysema quantification results using multivariate Cox proportional hazards models.

RESULTS

The study comprised 5178 participants (mean age ± SD, 66 ± 3 years; 3110 males). The median LAA-950 (18.2% vs 2.6%) and the proportion of LDCTs with LAA-950 exceeding 6% (96.3% vs 39.3%) saw a significant decline after kernel adaptation. There was no association between emphysema quantification before kernel adaptation and the risk of nonaccidental death. Nevertheless, after kernel adaptation, higher LAA-950 (hazards ratio for 1% increase, 1.01; P = 0.045) and LAA-950 exceeding 6% (hazards ratio, 1.36; P = 0.008) emerged as independent predictors of nonaccidental death, upon adjusting for age, sex, and smoking status.

CONCLUSIONS

The application of deep learning for kernel adaptation proves instrumental in quantifying pulmonary emphysema on LDCTs, establishing itself as a potential predictive tool for long-term nonaccidental mortality in asymptomatic individuals.

Diffusion histogram profiles predict molecular features of grade 4 in histologically lower-grade adult diffuse gliomas following WHO classification 2021

OBJECTIVES

In the latest World Health Organization classification 2021, grade 4 adult diffuse gliomas can be diagnosed with several molecular features even without histological evidence of necrosis or microvascular proliferation. We aimed to explore whole tumor histogram-derived apparent diffusion coefficient (ADC) histogram profiles for differentiating between the presence (Mol-4) and absence (Mol-2/3) of grade 4 molecular features in histologically lower-grade gliomas.

METHODS

Between June 2019 and October 2022, 184 adult patients with diffuse gliomas underwent MRI. After excluding 121 patients, 18 (median age, 64.5 [range, 37-84 years]) Mol-4 and 45 (median 40 [range, 18-73] years) Mol-2/3 patients with histologically lower-grade gliomas were enrolled. Whole tumor volume-of-interest-derived ADC histogram profiles were calculated and compared between the two groups. Stepwise logistic regression analysis with Akaike's information criterion using the ADC histogram profiles with p values < 0.01 and age at diagnosis was used to identify independent variables for predicting the Mol-4 group.

RESULTS

The 90th percentile (p < 0.001), median (p < 0.001), mean (p < 0.001), 10th percentile (p = 0.014), and entropy (p < 0.001) of normalized ADC were lower, and kurtosis (p < 0.001) and skewness (p = 0.046) were higher in the Mol-4 group than in the Mol-2/3 group. Multivariate logistic regression analysis revealed that the entropy of normalized ADC and age at diagnosis were independent predictive parameters for the Mol-4 group with an area under the curve of 0.92.

CONCLUSION

ADC histogram profiles may be promising preoperative imaging biomarkers to predict molecular grade 4 among histologically lower-grade adult diffuse gliomas.

CLINICAL RELEVANCE STATEMENT

This study highlighted the diagnostic usefulness of ADC histogram profiles to differentiate histologically lower grade adult diffuse gliomas with the presence of molecular grade 4 features and those without.

KEY POINTS

• ADC histogram profiles to predict molecular CNS WHO grade 4 status among histologically lower-grade adult diffuse gliomas were evaluated. • Entropy of ADC and age were independent predictive parameters for molecular grade 4 status. • ADC histogram analysis is useful for predicting molecular grade 4 among histologically lower-grade gliomas.

Comparing deep learning and handcrafted radiomics to predict chemoradiotherapy response for locally advanced cervical cancer using pretreatment MRI

Concurrent chemoradiotherapy (CRT) is the standard treatment for locally advanced cervical cancer (LACC), but its responsiveness varies among patients. A reliable tool for predicting CRT responses is necessary for personalized cancer treatment. In this study, we constructed prediction models using handcrafted radiomics (HCR) and deep learning radiomics (DLR) based on pretreatment MRI data to predict CRT response in LACC. Furthermore, we investigated the potential improvement in prediction performance by incorporating clinical factors. A total of 252 LACC patients undergoing curative chemoradiotherapy are included. The patients are randomly divided into two independent groups for the training (167 patients) and test datasets (85 patients). Contrast-enhanced T1- and T2-weighted MR scans are obtained. For HCR analysis, 1890 imaging features are extracted and a support vector machine classifier with a five-fold cross-validation is trained on training dataset to predict CRT response and subsequently validated on test dataset. For DLR analysis, a 3-dimensional convolutional neural network was trained on training dataset and validated on test dataset. In conclusion, both HCR and DLR models could predict CRT responses in patients with LACC. The integration of clinical factors into radiomics prediction models tended to improve performance in HCR analysis. Our findings may contribute to the development of personalized treatment strategies for LACC patients.

Development and multicenter validation of a multiparametric imaging model to predict treatment response in rectal cancer

OBJECTIVES

To develop and validate a multiparametric model to predict neoadjuvant treatment response in rectal cancer at baseline using a heterogeneous multicenter MRI dataset.

METHODS

Baseline staging MRIs (T2W (T2-weighted)-MRI, diffusion-weighted imaging (DWI) / apparent diffusion coefficient (ADC)) of 509 patients (9 centres) treated with neoadjuvant chemoradiotherapy (CRT) were collected. Response was defined as (1) complete versus incomplete response, or (2) good (Mandard tumor regression grade (TRG) 1-2) versus poor response (TRG3-5). Prediction models were developed using combinations of the following variable groups: (1) Non-imaging: age/sex/tumor-location/tumor-morphology/CRT-surgery interval (2) Basic staging: cT-stage/cN-stage/mesorectal fascia involvement, derived from (2a) original staging reports, or (2b) expert re-evaluation (3) Advanced staging: variables from 2b combined with cTN-substaging/invasion depth/extramural vascular invasion/tumor length (4) Quantitative imaging: tumour volume + first-order histogram features (from T2W-MRI and DWI/ADC) Models were developed with data from 6 centers (n = 412) using logistic regression with the Least Absolute Shrinkage and Selector Operator (LASSO) feature selection, internally validated using repeated (n = 100) random hold-out validation, and externally validated using data from 3 centers (n = 97).

RESULTS

After external validation, the best model (including non-imaging and advanced staging variables) achieved an area under the curve of 0.60 (95%CI=0.48-0.72) to predict complete response and 0.65 (95%CI=0.53-0.76) to predict a good response. Quantitative variables did not improve model performance. Basic staging variables consistently achieved lower performance compared to advanced staging variables.

CONCLUSIONS

Overall model performance was moderate. Best results were obtained using advanced staging variables, highlighting the importance of good-quality staging according to current guidelines. Quantitative imaging features had no added value (in this heterogeneous dataset).

CLINICAL RELEVANCE STATEMENT

Predicting tumour response at baseline could aid in tailoring neoadjuvant therapies for rectal cancer. This study shows that image-based prediction models are promising, though are negatively affected by variations in staging quality and MRI acquisition, urging the need for harmonization.

KEY POINTS

This multicenter study combining clinical information and features derived from MRI rendered disappointing performance to predict response to neoadjuvant treatment in rectal cancer. Best results were obtained with the combination of clinical baseline information and state-of-the-art image-based staging variables, highlighting the importance of good quality staging according to current guidelines and staging templates. No added value was found for quantitative imaging features in this multicenter retrospective study. This is likely related to acquisition variations, which is a major problem for feature reproducibility and thus model generalizability.

Artificial intelligence-based radiomics in bone tumors: Technical advances and clinical application

Radiomics is the extraction of predefined mathematic features from medical images for predicting variables of clinical interest. Recent research has demonstrated that radiomics can be processed by artificial intelligence algorithms to reveal complex patterns and trends for diagnosis, and prediction of prognosis and response to treatment modalities in various types of cancer. Artificial intelligence tools can utilize radiological images to solve next-generation issues in clinical decision making. Bone tumors can be classified as primary and secondary (metastatic) tumors. Osteosarcoma, Ewing sarcoma, and chondrosarcoma are the dominating primary tumors of bone. The development of bone tumor model systems and relevant research, and the assessment of novel treatment methods are ongoing to improve clinical outcomes, notably for patients with metastases. Artificial intelligence and radiomics have been utilized in almost full spectrum of clinical care of bone tumors. Radiomics models have achieved excellent performance in the diagnosis and grading of bone tumors. Furthermore, the models enable to predict overall survival, metastases, and recurrence. Radiomics features have exhibited promise in assisting therapeutic planning and evaluation, especially neoadjuvant chemotherapy. This review provides an overview of the evolution and opportunities for artificial intelligence in imaging, with a focus on hand-crafted features and deep learning-based radiomics approaches. We summarize the current application of artificial intelligence-based radiomics both in primary and metastatic bone tumors, and discuss the limitations and future opportunities of artificial intelligence-based radiomics in this field. In the era of personalized medicine, our in-depth understanding of emerging artificial intelligence-based radiomics approaches will bring innovative solutions to bone tumors and achieve clinical application.

The role of parametric feature maps to correct different volume of interest sizes: an in vivo liver MRI study

BACKGROUND

Different volume of interest (VOI) sizes influence radiomic features. This study examined if translating images into feature maps before feature sampling could compensate for these effects in liver magnetic resonance imaging (MRI).

METHODS

T1- and T2-weighted sequences from three different scanners (two 3-T scanners, one 1.5-T scanner) of 66 patients with normal abdominal MRI were included retrospectively. Three differently sized VOIs (10, 20, and 30 mm in diameter) were drawn in the liver parenchyma (right lobe), excluding adjacent structures. Ninety-three features were extracted conventionally using PyRadiomics. All images were also converted to 93 parametric feature maps using a pretested software. Agreement between the three VOI sizes was assessed with overall concordance correlation coefficients (OCCCs), while OCCCs > 0.85 were rated reproducible. OCCCs were calculated twice: for the VOI sizes of 10, 20, and 30 mm and for those of 20 and 30 mm.

RESULTS

When extracted from original images, only 4 out of the 93 features were reproducible across all VOI sizes in T1- and T2-weighted images. When the smallest VOI was excluded, 5 features (T1-weighted) and 7 features (T2-weighted) were reproducible. Extraction from parametric maps increased the number of reproducible features to 9 (T1- and T2-weighted) across all VOIs. Excluding the 10-mm VOI, reproducibility improved to 16 (T1-weighted) and 55 features (T2-weighted). The stability of all other features also increased in feature maps.

CONCLUSIONS

Translating images into parametric maps before feature extraction improves reproducibility across different VOI sizes in normal liver MRI.

RELEVANCE STATEMENT

The size of the segmented VOI influences the feature quantity of radiomics, while software-based conversion of images into parametric feature maps before feature sampling improves reproducibility across different VOI sizes in MRI of normal liver tissue.

KEY POINTS

• Parametric feature maps can compensate for different VOI sizes. • The effect seems dependent on the VOI sizes and the MRI sequence. • Feature maps can visualize features throughout the entire image stack.

Artificial intelligence in estimating fractional flow reserve: a systematic literature review of techniques

BACKGROUND

Fractional Flow Reserve (FFR) is the gold standard for the functional evaluation of coronary arteries, which is effective in selecting patients for revascularization, avoiding unnecessary procedures, and reducing treatment costs. However, its use is limited due to invasiveness, high cost, and complexity. Therefore, the non-invasive estimation of FFR using artificial intelligence (AI) methods is crucial.

OBJECTIVE

This study aimed to identify the AI techniques used for FFR estimation and to explore the features of the studies that applied AI techniques in FFR estimation.

METHODS

The present systematic review was conducted by searching five databases, PubMed, Scopus, Web of Science, IEEE, and Science Direct, based on the search strategy of each database.

RESULTS

Five hundred seventy-three articles were extracted, and by applying the inclusion and exclusion criteria, twenty-five were finally selected for review. The findings revealed that AI methods, including Machine Learning (ML) and Deep Learning (DL), have been used to estimate the FFR.

CONCLUSION

This study shows that AI methods can be used non-invasively to estimate FFR, which can help physicians diagnose and treat coronary artery occlusion and provide significant clinical performance for patients.

Multiparametric MRI: From Simultaneous Rapid Acquisition Methods and Analysis Techniques Using Scoring, Machine Learning, Radiomics, and Deep Learning to the Generation of Novel Metrics

ABSTRACT

With the recent advancements in rapid imaging methods, higher numbers of contrasts and quantitative parameters can be acquired in less and less time. Some acquisition models simultaneously obtain multiparametric images and quantitative maps to reduce scan times and avoid potential issues associated with the registration of different images. Multiparametric magnetic resonance imaging (MRI) has the potential to provide complementary information on a target lesion and thus overcome the limitations of individual techniques. In this review, we introduce methods to acquire multiparametric MRI data in a clinically feasible scan time with a particular focus on simultaneous acquisition techniques, and we discuss how multiparametric MRI data can be analyzed as a whole rather than each parameter separately. Such data analysis approaches include clinical scoring systems, machine learning, radiomics, and deep learning. Other techniques combine multiple images to create new quantitative maps associated with meaningful aspects of human biology. They include the magnetic resonance g-ratio, the inner to the outer diameter of a nerve fiber, and the aerobic glycolytic index, which captures the metabolic status of tumor tissues.

Machine learning predicts histologic type and grade of canine gliomas based on MRI texture analysis

Conventional MRI features of canine gliomas subtypes and grades significantly overlap. Texture analysis (TA) quantifies image texture based on spatial arrangement of pixel intensities. Machine learning (ML) models based on MRI-TA demonstrate high accuracy in predicting brain tumor types and grades in human medicine. The aim of this retrospective, diagnostic accuracy study was to investigate the accuracy of ML-based MRI-TA in predicting canine gliomas histologic types and grades. Dogs with histopathological diagnosis of intracranial glioma and available brain MRI were included. Tumors were manually segmented across their entire volume in enhancing part, non-enhancing part, and peri-tumoral vasogenic edema in T2-weighted (T2w), T1-weighted (T1w), FLAIR, and T1w postcontrast sequences. Texture features were extracted and fed into three ML classifiers. Classifiers' performance was assessed using a leave-one-out cross-validation approach. Multiclass and binary models were built to predict histologic types (oligodendroglioma vs. astrocytoma vs. oligoastrocytoma) and grades (high vs. low), respectively. Thirty-eight dogs with a total of 40 masses were included. Machine learning classifiers had an average accuracy of 77% for discriminating tumor types and of 75.6% for predicting high-grade gliomas. The support vector machine classifier had an accuracy of up to 94% for predicting tumor types and up to 87% for predicting high-grade gliomas. The most discriminative texture features of tumor types and grades appeared related to the peri-tumoral edema in T1w images and to the non-enhancing part of the tumor in T2w images, respectively. In conclusion, ML-based MRI-TA has the potential to discriminate intracranial canine gliomas types and grades.

Better Together: Data Harmonization and Cross-Study Analysis of Abdominal MRI Data From UK Biobank and the German National Cohort

OBJECTIVES

The UK Biobank (UKBB) and German National Cohort (NAKO) are among the largest cohort studies, capturing a wide range of health-related data from the general population, including comprehensive magnetic resonance imaging (MRI) examinations. The purpose of this study was to demonstrate how MRI data from these large-scale studies can be jointly analyzed and to derive comprehensive quantitative image-based phenotypes across the general adult population.

MATERIALS AND METHODS

Image-derived features of abdominal organs (volumes of liver, spleen, kidneys, and pancreas; volumes of kidney hilum adipose tissue; and fat fractions of liver and pancreas) were extracted from T1-weighted Dixon MRI data of 17,996 participants of UKBB and NAKO based on quality-controlled deep learning generated organ segmentations. To enable valid cross-study analysis, we first analyzed the data generating process using methods of causal discovery. We subsequently harmonized data from UKBB and NAKO using the ComBat approach for batch effect correction. We finally performed quantile regression on harmonized data across studies providing quantitative models for the variation of image-derived features stratified for sex and dependent on age, height, and weight.

RESULTS

Data from 8791 UKBB participants (49.9% female; age, 63 ± 7.5 years) and 9205 NAKO participants (49.1% female, age: 51.8 ± 11.4 years) were analyzed. Analysis of the data generating process revealed direct effects of age, sex, height, weight, and the data source (UKBB vs NAKO) on image-derived features. Correction of data source-related effects resulted in markedly improved alignment of image-derived features between UKBB and NAKO. Cross-study analysis on harmonized data revealed comprehensive quantitative models for the phenotypic variation of abdominal organs across the general adult population.

CONCLUSIONS

Cross-study analysis of MRI data from UKBB and NAKO as proposed in this work can be helpful for future joint data analyses across cohorts linking genetic, environmental, and behavioral risk factors to MRI-derived phenotypes and provide reference values for clinical diagnostics.

Assessment of artificial intelligence (AI) reporting methodology in glioma MRI studies using the Checklist for AI in Medical Imaging (CLAIM)

PURPOSE

The Checklist for Artificial Intelligence in Medical Imaging (CLAIM) is a recently released guideline designed for the optimal reporting methodology of artificial intelligence (AI) studies. Gliomas are the most common form of primary malignant brain tumour and numerous outcomes derived from AI algorithms such as grading, survival, treatment-related effects and molecular status have been reported. The aim of the study is to evaluate the AI reporting methodology for outcomes relating to gliomas in magnetic resonance imaging (MRI) using the CLAIM criteria.

METHODS

A literature search was performed on three databases pertaining to AI augmentation of glioma MRI, published between the start of 2018 and the end of 2021 RESULTS: A total of 4308 articles were identified and 138 articles remained after screening. These articles were categorised into four main AI tasks: grading (n= 44), predicting molecular status (n= 50), predicting survival (n= 25) and distinguishing true tumour progression from treatment-related effects (n= 10). The average CLAIM score was 20/42 (range: 10-31). Studies most consistently reported the scientific background and clinical role of their AI approach. Areas of improvement were identified in the reporting of data collection, data management, ground truth and validation of AI performance.

CONCLUSION

AI may be a means of producing high-accuracy results for certain tasks in glioma MRI; however, there remain issues with reporting quality. AI reporting guidelines may aid in a more reproducible and standardised approach to reporting and will aid in clinical integration.

A multi-reader comparison of normal-appearing white matter normalization techniques for perfusion and diffusion MRI in brain tumors

PURPOSE

There remains no consensus normal-appearing white matter (NAWM) normalization method to compute normalized relative cerebral blood volume (nrCBV) and apparent diffusion coefficient (nADC) in brain tumors. This reader study explored nrCBV and nADC differences using different NAWM normalization methods.

METHODS

Thirty-five newly diagnosed glioma patients were studied. For each patient, two readers created four NAWM regions of interests: (1) a single plane in the centrum semiovale (CSOp), (2) 3 spheres in the centrum semiovale (CSOs), (3) a single plane in the slice of the tumor center (TUMp), and (4) 3 spheres in the slice of the tumor center (TUMs). Readers repeated NAWM segmentations 1 month later. Differences in nrCBV and nADC of the FLAIR hyperintense tumor, inter-/intra-reader variability, and time to segment NAWM were assessed. As a validation step, the diagnostic performance of each method for IDH-status prediction was evaluated.

RESULTS

Both readers obtained significantly different nrCBV (P < .001), nADC (P < .001), and time to segment NAWM (P < .001) between the four normalization methods. nrCBV and nADC were significantly different between CSO and TUM methods, but not between planar and spherical methods in the same NAWM region. Broadly, CSO methods were quicker than TUM methods, and spherical methods were quicker than planar methods. For all normalization techniques, inter-reader reproducibility and intra-reader repeatability were excellent (intraclass correlation coefficient > 0.9), and the IDH-status predictive performance remained similar.

CONCLUSION

The selected NAWM region significantly impacts nrCBV and nADC values. CSO methods, particularly CSOs, may be preferred because of time reduction, similar reader variability, and similar diagnostic performance compared to TUM methods.

Development and validation of a deep learning signature for predicting lymph node metastasis in lung adenocarcinoma: comparison with radiomics signature and clinical-semantic model

OBJECTIVE

To develop and validate a deep learning (DL) signature for predicting lymph node (LN) metastasis in patients with lung adenocarcinoma.

METHODS

A total of 612 patients with pathologically-confirmed lung adenocarcinoma were retrospectively enrolled and were randomly divided into training cohort (n = 489) and internal validation cohort (n = 123). Besides, 108 patients were enrolled and constituted an independent test cohort (n = 108). Patients' clinical characteristics and CT semantic features were collected. The radiomics features were derived from contrast-enhanced CT images. The clinical-semantic model and radiomics signature were built to predict LN metastasis. Furthermore, Swin Transformer was adopted to develop a DL signature predictive of LN metastasis. Model performance was evaluated by area under the receiver operating characteristic curve (AUC), sensitivity, specificity, calibration curve, and decision curve analysis. The comparisons of AUC were conducted by the DeLong test.

RESULTS

The proposed DL signature yielded an AUC of 0.948-0.961 across all three cohorts, significantly superior to both clinical-semantic model and radiomics signature (all p < 0.05). The calibration curves show that DL signature predicted probabilities fit well the actual observed probabilities of LN metastasis. DL signature gained a higher net benefit than both clinical-semantic model and radiomics signature. The incorporation of radiomics signature or clinical-semantic risk predictors failed to reveal an incremental value over the DL signature.

CONCLUSIONS

The proposed DL signature based on Swin Transformer achieved a promising performance in predicting LN metastasis and could confer important information in noninvasive mediastinal LN staging and individualized therapeutic options.

KEY POINTS

• Accurate prediction for lymph node metastasis is crucial to formulate individualized therapeutic options for patients with lung adenocarcinoma. • The deep learning signature yielded an AUC of 0.948-0.961 across all three cohorts in predicting lymph node metastasis, superior to both radiomics signature and clinical-semantic model. • The incorporation of radiomics signature or clinical-semantic risk predictors into deep learning signature failed to reveal an incremental value over deep learning signature.

Influence of Image Processing on Radiomic Features From Magnetic Resonance Imaging

OBJECTIVE

Before implementing radiomics in routine clinical practice, comprehensive knowledge about the repeatability and reproducibility of radiomic features is required. The aim of this study was to systematically investigate the influence of image processing parameters on radiomic features from magnetic resonance imaging (MRI) in terms of feature values as well as test-retest repeatability.

MATERIALS AND METHODS

Utilizing a phantom consisting of 4 onions, 4 limes, 4 kiwifruits, and 4 apples, we acquired a test-retest dataset featuring 3 of the most commonly used MRI sequences on a 3 T scanner, namely, a T1-weighted, a T2-weighted, and a fluid-attenuated inversion recovery sequence, each at high and low resolution. After semiautomatic image segmentation, image processing with systematic variation of image processing parameters was performed, including spatial resampling, intensity discretization, and intensity rescaling. For each respective image processing setting, a total of 45 radiomic features were extracted, corresponding to the following 7 matrices/feature classes: conventional indices, histogram matrix, shape matrix, gray-level zone length matrix, gray-level run length matrix, neighboring gray-level dependence matrix, and gray-level cooccurrence matrix. Systematic differences of individual features between different resampling steps were assessed using 1-way analysis of variance with Tukey-type post hoc comparisons to adjust for multiple testing. Test-retest repeatability of radiomic features was measured using the concordance correlation coefficient, dynamic range, and intraclass correlation coefficient.

RESULTS

Image processing influenced radiological feature values. Regardless of the acquired sequence and feature class, significant differences ( P < 0.05) in feature values were found when the size of the resampled voxels was too large, that is, bigger than 3 mm. Almost all higher-order features depended strongly on intensity discretization. The effects of intensity rescaling were negligible except for some features derived from T1-weighted sequences. For all sequences, the percentage of repeatable features (concordance correlation coefficient and dynamic range ≥ 0.9) varied considerably depending on the image processing settings. The optimal image processing setting to achieve the highest percentage of stable features varied per sequence. Irrespective of image processing, the fluid-attenuated inversion recovery sequence in high-resolution overall yielded the highest number of stable features in comparison with the other sequences (89% vs 64%-78% for the respective optimal image processing settings). Across all sequences, the most repeatable features were generally obtained for a spatial resampling close to the originally acquired voxel size and an intensity discretization to at least 32 bins.

CONCLUSION

Variation of image processing parameters has a significant impact on the values of radiomic features as well as their repeatability. Furthermore, the optimal image processing parameters differ for each MRI sequence. Therefore, it is recommended that these processing parameters be determined in corresponding test-retest scans before clinical application. Extensive repeatability, reproducibility, and validation studies as well as standardization are required before quantitative image analysis and radiomics can be reliably translated into routine clinical care.

Common carotid segmentation in 18 F-sodium fluoride PET/CT scans: Head-to-head comparison of artificial intelligence-based and manual method

BACKGROUND

Carotid atherosclerosis is a major cause of stroke, traditionally diagnosed late. Positron emission tomography/computed tomography (PET/CT) with ¹⁸ F-sodium fluoride (NaF) detects arterial wall micro-calcification long before macro-calcification becomes detectable by ultrasound, CT or magnetic resonance imaging. However, manual PET/CT processing is time-consuming and requires experience. We compared a convolutional neural network (CNN) approach with manual segmentation of the common carotids.

METHODS

Segmentation in NaF-PET/CT scans of 29 healthy volunteers and 20 angina pectoris patients were compared for segmented volume (Vol) and mean, maximal, and total standardized uptake values (SUVmean, SUVmax, and SUVtotal). SUVmean was the average of SUVmeans within the VOI, SUVmax the highest SUV in all voxels in the VOI, and SUVtotal the SUVmean multiplied by the Vol of the VOI. Intra and Interobserver variability with manual segmentation was examined in 25 randomly selected scans.

RESULTS

Bias for Vol, SUVmean, SUVmax, and SUVtotal were 1.33 ± 2.06, -0.01 ± 0.05, 0.09 ± 0.48, and 1.18 ± 1.99 in the left and 1.89 ± 1.5, -0.07 ± 0.12, 0.05 ± 0.47, and 1.61 ± 1.47, respectively, in the right common carotid artery. Manual segmentation lasted typically 20 min versus 1 min with the CNN-based approach. Mean Vol deviation at repeat manual segmentation was 14% and 27% in left and right common carotids.

CONCLUSIONS

CNN-based segmentation was much faster and provided SUVmean values virtually identical to manually obtained ones, suggesting CNN-based analysis as a promising substitute of slow and cumbersome manual processing.

Machine Learning Based on Computed Tomography Pulmonary Angiography in Evaluating Pulmonary Artery Pressure in Patients with Pulmonary Hypertension

BACKGROUND

Right heart catheterization is the gold standard for evaluating hemodynamic parameters of pulmonary circulation, especially pulmonary artery pressure (PAP) for diagnosis of pulmonary hypertension (PH). However, the invasive and costly nature of RHC limits its widespread application in daily practice.

PURPOSE

To develop a fully automatic framework for PAP assessment via machine learning based on computed tomography pulmonary angiography (CTPA).

MATERIALS AND METHODS

A machine learning model was developed to automatically extract morphological features of pulmonary artery and the heart on CTPA cases collected between June 2017 and July 2021 based on a single center experience. Patients with PH received CTPA and RHC examinations within 1 week. The eight substructures of pulmonary artery and heart were automatically segmented through our proposed segmentation framework. Eighty percent of patients were used for the training data set and twenty percent for the independent testing data set. PAP parameters, including mPAP, sPAP, dPAP, and TPR, were defined as ground-truth. A regression model was built to predict PAP parameters and a classification model to separate patients through mPAP and sPAP with cut-off values of 40 mm Hg and 55 mm Hg in PH patients, respectively. The performances of the regression model and the classification model were evaluated by analyzing the intraclass correlation coefficient (ICC) and the area under the receiver operating characteristic curve (AUC).

RESULTS

Study participants included 55 patients with PH (men 13; age 47.75 ± 14.87 years). The average dice score for segmentation increased from 87.3% ± 2.9 to 88.2% ± 2.9 through proposed segmentation framework. After features extraction, some of the AI automatic extractions (AAd, RVd, LAd, and RPAd) achieved good consistency with the manual measurements. The differences between them were not statistically significant (t = 1.222, p = 0.227; t = -0.347, p = 0.730; t = 0.484, p = 0.630; t = -0.320, p = 0.750, respectively). The Spearman test was used to find key features which are highly correlated with PAP parameters. Correlations between pulmonary artery pressure and CTPA features show a high correlation between mPAP and LAd, LVd, LAa (r = 0.333, p = 0.012; r = -0.400, p = 0.002; r = -0.208, p = 0.123; r = -0.470, p = 0.000; respectively). The ICC between the output of the regression model and the ground-truth from RHC of mPAP, sPAP, and dPAP were 0.934, 0.903, and 0.981, respectively. The AUC of the receiver operating characteristic curve of the classification model of mPAP and sPAP were 0.911 and 0.833.

CONCLUSIONS

The proposed machine learning framework on CTPA enables accurate segmentation of pulmonary artery and heart and automatic assessment of the PAP parameters and has the ability to accurately distinguish different PH patients with mPAP and sPAP. Results of this study may provide additional risk stratification indicators in the future with non-invasive CTPA data.

The Low Rate of Adherence to Checklist for Artificial Intelligence in Medical Imaging Criteria Among Published Prostate MRI Artificial Intelligence Algorithms

OBJECTIVE

To determine the rigor, generalizability, and reproducibility of published classification and detection artificial intelligence (AI) models for prostate cancer (PCa) on MRI using the Checklist for Artificial Intelligence in Medical Imaging (CLAIM) guidelines, a 42-item checklist that is considered a measure of best practice for presenting and reviewing medical imaging AI research.

MATERIALS AND METHODS

This review searched English literature for studies proposing PCa AI detection and classification models on MRI. Each study was evaluated with the CLAIM checklist. The additional outcomes for which data were sought included measures of AI model performance (eg, area under the curve [AUC], sensitivity, specificity, free-response operating characteristic curves), training and validation and testing group sample size, AI approach, detection versus classification AI, public data set utilization, MRI sequences used, and definition of gold standard for ground truth. The percentage of CLAIM checklist fulfillment was used to stratify studies into quartiles. Wilcoxon's rank-sum test was used for pair-wise comparisons.

RESULTS

In all, 75 studies were identified, and 53 studies qualified for analysis. The original CLAIM items that most studies did not fulfill includes item 12 (77% no): de-identification methods; item 13 (68% no): handling missing data; item 15 (47% no): rationale for choosing ground truth reference standard; item 18 (55% no): measurements of inter- and intrareader variability; item 31 (60% no): inclusion of validated interpretability maps; item 37 (92% no): inclusion of failure analysis to elucidate AI model weaknesses. An AUC score versus percentage CLAIM fulfillment quartile revealed a significant difference of the mean AUC scores between quartile 1 versus quartile 2 (0.78 versus 0.86, P = .034) and quartile 1 versus quartile 4 (0.78 versus 0.89, P = .003) scores. Based on additional information and outcome metrics gathered in this study, additional measures of best practice are defined. These new items include disclosure of public dataset usage, ground truth definition in comparison to other referenced works in the defined task, and sample size power calculation.

CONCLUSION

A large proportion of AI studies do not fulfill key items in CLAIM guidelines within their methods and results sections. The percentage of CLAIM checklist fulfillment is weakly associated with improved AI model performance. Additions or supplementations to CLAIM are recommended to improve publishing standards and aid reviewers in determining study rigor.

AAPM task group report 273: Recommendations on best practices for AI and machine learning for computer-aided diagnosis in medical imaging

Rapid advances in artificial intelligence (AI) and machine learning, and specifically in deep learning (DL) techniques, have enabled broad application of these methods in health care. The promise of the DL approach has spurred further interest in computer-aided diagnosis (CAD) development and applications using both "traditional" machine learning methods and newer DL-based methods. We use the term CAD-AI to refer to this expanded clinical decision support environment that uses traditional and DL-based AI methods. Numerous studies have been published to date on the development of machine learning tools for computer-aided, or AI-assisted, clinical tasks. However, most of these machine learning models are not ready for clinical deployment. It is of paramount importance to ensure that a clinical decision support tool undergoes proper training and rigorous validation of its generalizability and robustness before adoption for patient care in the clinic. To address these important issues, the American Association of Physicists in Medicine (AAPM) Computer-Aided Image Analysis Subcommittee (CADSC) is charged, in part, to develop recommendations on practices and standards for the development and performance assessment of computer-aided decision support systems. The committee has previously published two opinion papers on the evaluation of CAD systems and issues associated with user training and quality assurance of these systems in the clinic. With machine learning techniques continuing to evolve and CAD applications expanding to new stages of the patient care process, the current task group report considers the broader issues common to the development of most, if not all, CAD-AI applications and their translation from the bench to the clinic. The goal is to bring attention to the proper training and validation of machine learning algorithms that may improve their generalizability and reliability and accelerate the adoption of CAD-AI systems for clinical decision support.

Harmonization Strategies in Multicenter MRI-Based Radiomics

Radiomics analysis is a powerful tool aiming to provide diagnostic and prognostic patient information directly from images that are decoded into handcrafted features, comprising descriptors of shape, size and textural patterns. Although radiomics is gaining momentum since it holds great promise for accelerating digital diagnostics, it is susceptible to bias and variation due to numerous inter-patient factors (e.g., patient age and gender) as well as inter-scanner ones (different protocol acquisition depending on the scanner center). A variety of image and feature based harmonization methods has been developed to compensate for these effects; however, to the best of our knowledge, none of these techniques has been established as the most effective in the analysis pipeline so far. To this end, this review provides an overview of the challenges in optimizing radiomics analysis, and a concise summary of the most relevant harmonization techniques, aiming to provide a thorough guide to the radiomics harmonization process.

Artificial intelligence and machine learning in cancer imaging

An increasing array of tools is being developed using artificial intelligence (AI) and machine learning (ML) for cancer imaging. The development of an optimal tool requires multidisciplinary engagement to ensure that the appropriate use case is met, as well as to undertake robust development and testing prior to its adoption into healthcare systems. This multidisciplinary review highlights key developments in the field. We discuss the challenges and opportunities of AI and ML in cancer imaging; considerations for the development of algorithms into tools that can be widely used and disseminated; and the development of the ecosystem needed to promote growth of AI and ML in cancer imaging.

Artificial intelligence in cardiac magnetic resonance fingerprinting

Magnetic resonance fingerprinting (MRF) is a fast MRI-based technique that allows for multiparametric quantitative characterization of the tissues of interest in a single acquisition. In particular, it has gained attention in the field of cardiac imaging due to its ability to provide simultaneous and co-registered myocardial T₁ and T₂ mapping in a single breath-held cardiac MRF scan, in addition to other parameters. Initial results in small healthy subject groups and clinical studies have demonstrated the feasibility and potential of MRF imaging. Ongoing research is being conducted to improve the accuracy, efficiency, and robustness of cardiac MRF. However, these improvements usually increase the complexity of image reconstruction and dictionary generation and introduce the need for sequence optimization. Each of these steps increase the computational demand and processing time of MRF. The latest advances in artificial intelligence (AI), including progress in deep learning and the development of neural networks for MRI, now present an opportunity to efficiently address these issues. Artificial intelligence can be used to optimize candidate sequences and reduce the memory demand and computational time required for reconstruction and post-processing. Recently, proposed machine learning-based approaches have been shown to reduce dictionary generation and reconstruction times by several orders of magnitude. Such applications of AI should help to remove these bottlenecks and speed up cardiac MRF, improving its practical utility and allowing for its potential inclusion in clinical routine. This review aims to summarize the latest developments in artificial intelligence applied to cardiac MRF. Particularly, we focus on the application of machine learning at different steps of the MRF process, such as sequence optimization, dictionary generation and image reconstruction.

Enhancing the stability of CT radiomics across different volume of interest sizes using parametric feature maps: a phantom study

BACKGROUND

In radiomics studies, differences in the volume of interest (VOI) are often inevitable and may confound the extracted features. We aimed to correct this confounding effect of VOI variability by applying parametric maps with a fixed voxel size.

METHODS

Ten scans of a cup filled with sodium chloride solution were scanned using a multislice computed tomography (CT) unit. Sphere-shaped VOIs with different diameters (4, 8, or 16 mm) were drawn centrally into the phantom. A total of 93 features were extracted conventionally from the original images using PyRadiomics. Using a self-designed and pretested software tool, parametric maps for the same 93 features with a fixed voxel size of 4 mm3 were created. To retrieve the feature values from the maps, VOIs were copied from the original images to preserve the position. Differences in feature quantities between the VOI sizes were tested with the Mann-Whitney U-test and agreement with overall concordance correlation coefficients (OCCC).

RESULTS

Fifty-five conventionally extracted features were significantly different between the VOI sizes, and none of the features showed excellent agreement in terms of OCCCs. When read from the parametric maps, only 8 features showed significant differences, and 3 features showed an excellent OCCC (≥ 0.85). The OCCCs for 89 features substantially increased using the parametric maps.

CONCLUSIONS

This phantom study shows that converting CT images into parametric maps resolves the confounding effect of VOI variability and increases feature reproducibility across VOI sizes.

Pseudoprospective Paraclinical Interaction of Radiology Residents With a Deep Learning System for Prostate Cancer Detection: Experience, Performance, and Identification of the Need for Intermittent Recalibration

OBJECTIVES

The aim of this study was to estimate the prospective utility of a previously retrospectively validated convolutional neural network (CNN) for prostate cancer (PC) detection on prostate magnetic resonance imaging (MRI).

MATERIALS AND METHODS

The biparametric (T2-weighted and diffusion-weighted) portion of clinical multiparametric prostate MRI from consecutive men included between November 2019 and September 2020 was fully automatically and individually analyzed by a CNN briefly after image acquisition (pseudoprospective design). Radiology residents performed 2 research Prostate Imaging Reporting and Data System (PI-RADS) assessments of the multiparametric dataset independent from clinical reporting (paraclinical design) before and after review of the CNN results and completed a survey. Presence of clinically significant PC was determined by the presence of an International Society of Urological Pathology grade 2 or higher PC on combined targeted and extended systematic transperineal MRI/transrectal ultrasound fusion biopsy. Sensitivities and specificities on a patient and prostate sextant basis were compared using the McNemar test and compared with the receiver operating characteristic (ROC) curve of CNN. Survey results were summarized as absolute counts and percentages.

RESULTS

A total of 201 men were included. The CNN achieved an ROC area under the curve of 0.77 on a patient basis. Using PI-RADS ≥3-emulating probability threshold (c3), CNN had a patient-based sensitivity of 81.8% and specificity of 54.8%, not statistically different from the current clinical routine PI-RADS ≥4 assessment at 90.9% and 54.8%, respectively ( P = 0.30/ P = 1.0). In general, residents achieved similar sensitivity and specificity before and after CNN review. On a prostate sextant basis, clinical assessment possessed the highest ROC area under the curve of 0.82, higher than CNN (AUC = 0.76, P = 0.21) and significantly higher than resident performance before and after CNN review (AUC = 0.76 / 0.76, P ≤ 0.03). The resident survey indicated CNN to be helpful and clinically useful.

CONCLUSIONS

Pseudoprospective paraclinical integration of fully automated CNN-based detection of suspicious lesions on prostate multiparametric MRI was demonstrated and showed good acceptance among residents, whereas no significant improvement in resident performance was found. General CNN performance was preserved despite an observed shift in CNN calibration, identifying the requirement for continuous quality control and recalibration.

Statistical considerations for repeatability and reproducibility of quantitative imaging biomarkers

Quantitative imaging biomarkers (QIBs) are increasingly used in clinical studies. Because many QIBs are derived through multiple steps in image data acquisition and data analysis, QIB measurements can produce large variabilities, posing a significant challenge in translating QIBs into clinical trials, and ultimately, clinical practice. Both repeatability and reproducibility constitute the reliability of a QIB measurement. In this article, we review the statistical aspects of repeatability and reproducibility of QIB measurements by introducing methods and metrics for assessments of QIB repeatability and reproducibility and illustrating the impact of QIB measurement error on sample size and statistical power calculations, as well as predictive performance with a QIB as a predictive biomarker.

Short-term repeatability and long-term reproducibility of quantitative MR imaging biomarkers in a single centre longitudinal study

Quantitative imaging biomarkers (QIBs) can be defined as objective measures that are sensitive and specific to changes in tissue physiology. Provided the acquired QIBs are not affected by scanner changes, they could play an important role in disease diagnosis, prognosis, management, and treatment monitoring. The precision of selected QIBs was assessed from data collected on a 3-T scanner in four healthy participants over a 5-year period. Inevitable scanner changes and acquisition protocol revisions occurred during this time. Standard and custom processing pipelines were used to calculate regional brain volume, cortical thickness, T2, T2*, quantitative susceptibility, cerebral blood flow, axial, radial and mean diffusivity, peak width of skeletonized mean diffusivity, and fractional anisotropy from the acquired images. Coefficient of variation (CoV) and intra-class correlation (ICC) indices were determined in the short-term (i.e., repeatable over three acquisitions within 4 weeks) and in the long-term (i.e., reproducible over four acquisition sessions in 5 years). Precision indices varied based on acquisition technique, processing pipeline, and anatomical region. Good repeatability (average CoV=2.40% and ICC=0.78) and reproducibility (average CoV=8.86 % and ICC=0.72) were found over all QIBs. The best performance indices were obtained for diffusion derived biomarkers (CoV∼0.96% and ICCs=0.87); conversely, the poorest indices were found for the cerebral blood flow biomarker (CoV>10% and ICC<0.5). These results demonstrate that changes in protocol, along with hardware and software upgrades, did not affect the estimates of the selected biomarkers and their precision. Further characterization of the QIB is necessary to understand meaningful changes in the biomarkers in longitudinal studies of normal brain aging and translation to clinical research.

The Compressed Sensing MP2RAGE as a Surrogate to the MPRAGE for Neuroimaging at 3 T

OBJECTIVES

The magnetization-prepared 2 rapid acquisition gradient echo (MP2RAGE) sequence provides quantitative T1 maps in addition to high-contrast morphological images. Advanced acceleration techniques such as compressed sensing (CS) allow its acquisition time to be compatible with clinical applications. To consider its routine use in future neuroimaging protocols, the repeatability of the segmented brain structures was evaluated and compared with the standard morphological sequence (magnetization-prepared rapid gradient echo [MPRAGE]). The repeatability of the T1 measurements was also assessed.

MATERIALS AND METHODS

Thirteen healthy volunteers were scanned either 3 or 4 times at several days of interval, on a 3 T clinical scanner, with the 2 sequences (CS-MP2RAGE and MPRAGE), set with the same spatial resolution (0.8-mm isotropic) and scan duration (6 minutes 21 seconds). The reconstruction time of the CS-MP2RAGE outputs (including the 2 echo images, the MP2RAGE image, and the T1 map) was 3 minutes 33 seconds, using an open-source in-house algorithm implemented in the Gadgetron framework.Both precision and variability of volume measurements obtained from CAT12 and VolBrain were assessed. The T1 accuracy and repeatability were measured on phantoms and on humans and were compared with literature.Volumes obtained from the CS-MP2RAGE and the MPRAGE images were compared using Student t tests (P < 0.05 was considered significant).

RESULTS

The CS-MP2RAGE acquisition provided morphological images of the same quality and higher contrasts than the standard MPRAGE images. Similar intravolunteer variabilities were obtained with the CS-MP2RAGE and the MPRAGE segmentations. In addition, high-resolution T1 maps were obtained from the CS-MP2RAGE. T1 times of white and gray matters and several deep gray nuclei are consistent with the literature and show very low variability (<1%).

CONCLUSIONS

The CS-MP2RAGE can be used in future protocols to rapidly obtain morphological images and quantitative T1 maps in 3-dimensions while maintaining high repeatability in volumetry and relaxation times.

Logistic Regression-Based Model Is More Efficient Than U-Net Model for Reliable Whole Brain Magnetic Resonance Imaging Segmentation

OBJECTIVES

Automated whole brain segmentation from magnetic resonance images is of great interest for the development of clinically relevant volumetric markers for various neurological diseases. Although deep learning methods have demonstrated remarkable potential in this area, they may perform poorly in nonoptimal conditions, such as limited training data availability. Manual whole brain segmentation is an incredibly tedious process, so minimizing the data set size required for training segmentation algorithms may be of wide interest. The purpose of this study was to compare the performance of the prototypical deep learning segmentation architecture (U-Net) with a previously published atlas-free traditional machine learning method, Classification using Derivative-based Features (C-DEF) for whole brain segmentation, in the setting of limited training data.

MATERIALS AND METHODS

C-DEF and U-Net models were evaluated after training on manually curated data from 5, 10, and 15 participants in 2 research cohorts: (1) people living with clinically diagnosed HIV infection and (2) relapsing-remitting multiple sclerosis, each acquired at separate institutions, and between 5 and 295 participants' data using a large, publicly available, and annotated data set of glioblastoma and lower grade glioma (brain tumor segmentation). Statistics was performed on the Dice similarity coefficient using repeated-measures analysis of variance and Dunnett-Hsu pairwise comparison.

RESULTS

C-DEF produced better segmentation than U-Net in lesion (29.2%-38.9%) and cerebrospinal fluid (5.3%-11.9%) classes when trained with data from 15 or fewer participants. Unlike C-DEF, U-Net showed significant improvement when increasing the size of the training data (24%-30% higher than baseline). In the brain tumor segmentation data set, C-DEF produced equivalent or better segmentations than U-Net for enhancing tumor and peritumoral edema regions across all training data sizes explored. However, U-Net was more effective than C-DEF for segmentation of necrotic/non-enhancing tumor when trained on 10 or more participants, probably because of the inconsistent signal intensity of the tissue class.

CONCLUSIONS

These results demonstrate that classical machine learning methods can produce more accurate brain segmentation than the far more complex deep learning methods when only small or moderate amounts of training data are available (n ≤ 15). The magnitude of this advantage varies by tissue and cohort, while U-Net may be preferable for deep gray matter and necrotic/non-enhancing tumor segmentation, particularly with larger training data sets (n ≥ 20). Given that segmentation models often need to be retrained for application to novel imaging protocols or pathology, the bottleneck associated with large-scale manual annotation could be avoided with classical machine learning algorithms, such as C-DEF.

Plant Genotype to Phenotype Prediction Using Machine Learning

Genomic prediction tools support crop breeding based on statistical methods, such as the genomic best linear unbiased prediction (GBLUP). However, these tools are not designed to capture non-linear relationships within multi-dimensional datasets, or deal with high dimension datasets such as imagery collected by unmanned aerial vehicles. Machine learning (ML) algorithms have the potential to surpass the prediction accuracy of current tools used for genotype to phenotype prediction, due to their capacity to autonomously extract data features and represent their relationships at multiple levels of abstraction. This review addresses the challenges of applying statistical and machine learning methods for predicting phenotypic traits based on genetic markers, environment data, and imagery for crop breeding. We present the advantages and disadvantages of explainable model structures, discuss the potential of machine learning models for genotype to phenotype prediction in crop breeding, and the challenges, including the scarcity of high-quality datasets, inconsistent metadata annotation and the requirements of ML models.

The Road Toward Reproducibility of Parametric Mapping of the Heart: A Technical Review

Parametric mapping of the heart has become an essential part of many cardiovascular magnetic resonance imaging exams, and is used for tissue characterization and diagnosis in a broad range of cardiovascular diseases. These pulse sequences are used to quantify the myocardial T₁, T₂, T2*, and T_1ρ relaxation times, which are unique surrogate indices of fibrosis, edema and iron deposition that can be used to monitor a disease over time or to compare patients to one another. Parametric mapping is now well-accepted in the clinical setting, but its wider dissemination is hindered by limited inter-center reproducibility and relatively long acquisition times. Recently, several new parametric mapping techniques have appeared that address both of these problems, but substantial hurdles remain for widespread clinical adoption. This review serves both as a primer for newcomers to the field of parametric mapping and as a technical update for those already well at home in it. It aims to establish what is currently needed to improve the reproducibility of parametric mapping of the heart. To this end, we first give an overview of the metrics by which a mapping technique can be assessed, such as bias and variability, as well as the basic physics behind the relaxation times themselves and what their relevance is in the prospect of myocardial tissue characterization. This is followed by a summary of routine mapping techniques and their variations. The problems in reproducibility and the sources of bias and variability of these techniques are reviewed. Subsequently, novel fast, whole-heart, and multi-parametric techniques and their merits are treated in the light of their reproducibility. This includes state of the art segmentation techniques applied to parametric maps, and how artificial intelligence is being harnessed to solve this long-standing conundrum. We finish up by sketching an outlook on the road toward inter-center reproducibility, and what to expect in the future.

Association between Chest CT-defined Emphysema and Lung Cancer: A Systematic Review and Meta-Analysis

Background Given the different methods of assessing emphysema, controversy exists as to whether it is associated with lung cancer. Purpose To perform a systematic review and meta-analysis of the association between chest CT-defined emphysema and the presence of lung cancer. Materials and Methods The PubMed, Embase, and Cochrane databases were searched up to July 15, 2021, to identify studies on the association between emphysema assessed visually or quantitatively with CT and lung cancer. Associations were determined by emphysema severity (trace, mild, or moderate to severe, assessed visually and quantitatively) and subtype (centrilobular and paraseptal, assessed visually). Overall and stratified pooled odds ratios (ORs) with their 95% CIs were obtained. Results Of the 3343 screened studies, 21 studies (107 082 patients) with 26 subsets were included. The overall pooled ORs for lung cancer given the presence of emphysema were 2.3 (95% CI: 2.0, 2.6; I² = 35%; 19 subsets) and 1.02 (95% CI: 1.01, 1.02; six subsets) per 1% increase in low attenuation area. Studies with visual (pooled OR, 2.3; 95% CI: 1.9, 2.6; I² = 48%; 12 subsets) and quantitative (pooled OR, 2.2; 95% CI: 1.8, 2.8; I² = 3.7%; eight subsets) assessments yielded comparable results for the dichotomous assessment. Based on six studies (1716 patients), the pooled ORs for lung cancer increased with emphysema severity and were higher for visual assessment (2.5, 3.7, and 4.5 for trace, mild, and moderate to severe, respectively) than for quantitative assessment (1.9, 2.2, and 2.5) based on point estimates. Compared with no emphysema, only centrilobular emphysema (three studies) was associated with lung cancer (pooled OR, 2.2; 95% CI: 1.5, 3.2; P < .001). Conclusion Both visual and quantitative CT assessments of emphysema were associated with a higher odds of lung cancer, which also increased with emphysema severity. Regarding subtype, only centrilobular emphysema was significantly associated with lung cancer. Clinical trial registration no. CRD42021262163 © RSNA, 2022 See also the editorial by Hunsaker in this issue. Online supplemental material is available for this article.

Radiopaque Recreations of Lung Pathologies From Clinical Computed Tomography Images Using Potassium Iodide Inkjet 3-dimensional Printing: Proof of Concept

PURPOSE

The purpose of this study was to develop a 3-dimensional (3D) printing method to create computed tomography (CT) realistic phantoms of lung cancer nodules and lung parenchymal disease from clinical CT images.

MATERIALS AND METHODS

Low-density paper was used as substrate material for inkjet printing with potassium iodide solution to reproduce phantoms that mimic the CT attenuation of lung parenchyma. The relationship between grayscale values and the corresponding CT numbers of prints was first established through the derivation of exponential fitted equation from scanning data. Next, chest CTs from patients with early-stage lung cancer and coronavirus disease 2019 (COVID-19) pneumonia were chosen for 3D printing. CT images of original lung nodule and the 3D-printed nodule phantom were compared based on pixel-to-pixel correlation and radiomic features.

RESULTS

CT images of part-solid lung cancer and 3D-printed nodule phantom showed both high visual similarity and quantitative correlation. R2 values from linear regressions of pixel-to-pixel correlations between 5 sets of patient and 3D-printed image pairs were 0.92, 0.94, 0.86, 0.85, and 0.83, respectively. Comparison of radiomic measures between clinical CT and printed models demonstrated 6.1% median difference, with 25th and 75th percentile range at 2.4% and 15.2% absolute difference, respectively. The densities and parenchymal morphologies from COVID-19 pneumonia CT images were well reproduced in the 3D-printed phantom scans.

CONCLUSION

The 3D printing method presented in this work facilitates creation of CT-realistic reproductions of lung cancer and parenchymal disease from individual patient scans with microbiological and pathology confirmation.

Deep Learning-Based Image Conversion Improves the Reproducibility of Computed Tomography Radiomics Features: A Phantom Study

OBJECTIVES

This study aimed to evaluate the usefulness of deep learning-based image conversion to improve the reproducibility of computed tomography (CT) radiomics features.

MATERIALS AND METHODS

This study was conducted using an abdominal phantom with liver nodules. We developed an image conversion algorithm using a residual feature aggregation network to reproduce radiomics features with CT images under various CT protocols and reconstruction kernels. External validation was performed using images from different scanners, consisting of 8 different protocols. To evaluate the variability of radiomics features, regions of interest (ROIs) were drawn by targeting the liver parenchyma, vessels, paraspinal area, and liver nodules. We extracted 18 first-order, 68 second-order, and 688 wavelet radiomics features. Measurement variability was assessed using the concordance correlation coefficient (CCC), compared with the ground-truth image.

RESULTS

In the ROI-based analysis, there was an 83.3% improvement of CCC (80/96; 4 ROIs with 3 categories of radiomics features and 8 protocols) in synthetic images compared with the original images. Among them, the 56 CCC pairs showed a significant increase after image synthesis. In the radiomics feature-based analysis, 62.0% (3838 of 6192; 774 radiomics features with 8 protocols) features showed increased CCC after image synthesis, and a significant increase was noted in 26.9% (1663 of 6192) features. In particular, the first-order feature (79.9%, 115/144) showed better improvement in terms of the reproducibility of radiomics feature than the second-order (59.9%, 326/544) or wavelet feature (61.7%, 3397/5504).

CONCLUSIONS

Our study demonstrated that a deep learning model for image conversion can improve the reproducibility of radiomics features across various CT protocols, reconstruction kernels, and CT scanners.

Dual-Echo Turbo Spin Echo and 12-Echo Multi Spin Echo Sequences as Equivalent Techniques for Obtaining T2-Relaxometry Data: Application in Symptomatic and Asymptomatic Hereditary Transthyretin Amyloidosis as a Surrogate Disease

OBJECTIVES

Multi spin echo (MSE) sequences are often used for obtaining T2-relaxometry data as they provide defined echo times (TEs). Due to their time-consuming acquisition, they are frequently replaced by turbo spin echo (TSE) sequences that in turn bear the risk of systematic errors when analyzing small structures or lesions. With this study, we aim to test whether T2-relaxometry data derived from either dual-echo TSE or 12-echo MSE sequences are equivalent for quantifying peripheral nerve lesions. Hereditary transthyretin (ATTRv) amyloidosis was chosen as a surrogate disease, as it allows the inclusion of both asymptomatic carriers of the underlying variant transthyretin gene (varTTR) and symptomatic ATTRv amyloidosis patients.

MATERIALS AND METHODS

Overall, 50 participants with genetically confirmed varTTR (20 clinically symptomatic ATTRv amyloidosis; 4 females, 16 males; mean age, 61.8 years; range, 33-76 years; and 30 asymptomatic varTTR-carriers; 18 females, 12 males; mean age, 43.1 years; range, 21-62 years), and 30 healthy volunteers (13 females, 17 males, mean age 41.3 years, range 22-73) were prospectively included and underwent magnetic resonance neurography at 3 T. T2-relaxometry was performed by acquiring an axial 2-dimensional dual-echo TSE sequence with spectral fat saturation (TE1/TE2, 12/73 milliseconds; TR, 5210 milliseconds; acquisition time, 7 minutes, 30 seconds), and an axial 2-dimensional MSE sequence with spectral fat saturation and with 12 different TE (TE1, 10 milliseconds to TE12, 120 milliseconds; ΔTE, 10 milliseconds; TR, 3000 milliseconds; acquisition time, 11 minutes, 23 seconds) at the right mid to lower thigh. Sciatic nerve regions of interest were manually drawn in ImageJ on 10 central slices per participant and sequence, and the apparent T2-relaxation time (T2app) and proton spin density (ρ) were calculated individually from TSE and MSE relaxometry data.

RESULTS

Linear regression showed that T2app values obtained from the dual-echo TSE (T2appTSE), and those calculated from the 12-echo MSE (T2appMSE) were mathematically connected by a factor of 1.3 throughout all groups (controls: 1.26 ± 0.02; varTTR-carriers: 1.25 ± 0.02; symptomatic ATTRv amyloidosis: 1.28 ± 0.02), whereas a factor of 0.5 was identified between respective ρ values (controls: 0.47 ± 0.01; varTTR-carriers: 0.47 ± 0.01; symptomatic ATTRv amyloidosis: 0.50 ± 0.02). T2app calculated from both TSE and MSE, distinguished between symptomatic ATTRv (T2appTSE 66.38 ± 2.6; T2appMSE 84.6 ± 3.3) and controls (T2appTSE 58.1 ± 1.0, P = 0.0028; T2appMSE 72.8 ± 0.7, P < 0.0001), whereas differences between varTTR-carriers (T2appTSE 61.8 ± 1.5; T2appMSE 76.7 ± 1.3) and ATTRv amyloidosis were observed only for T2appMSE (P = 0.0082). The ρ value differentiated well between healthy controls (ρTSE 365.1 ± 7.2; ρMSE 170.4 ± 3.8) versus varTTR-carriers (ρTSE 415.7 ± 9.8, P = 0.0027; ρMSE 193.7 ± 5.3, P = 0.0398) and versus symptomatic ATTRv amyloidosis (ρTSE 487.8 ± 17.9; ρMSE 244.7 ± 13.1, P < 0.0001, respectively), but also between varTTR-carriers and ATTRv amyloidosis (ρTSEP = 0.0001; ρMSEP < 0.0001).

CONCLUSIONS

Dual-echo TSE and 12-echo MSE sequences provide equally robust and reliable T2-relaxometry data when calculating T2app and ρ. Due to their shorter acquisition time and higher resolution, TSE sequences may be preferred in future magnetic resonance imaging protocols. As a secondary result, ρ can be confirmed as a sensitive biomarker to detect early nerve lesions as it differentiated best among healthy controls, asymptomatic varTTR-carriers, and symptomatic ATTRv amyloidosis, whereas T2app might be beneficial in already manifest ATTRv amyloidosis.

Cycle-consistent adversarial networks improves generalizability of radiomics model in grading meningiomas on external validation

The heterogeneity of MRI is one of the major reasons for decreased performance of a radiomics model on external validation, limiting the model’s generalizability and clinical application. We aimed to establish a generalizable radiomics model to predict meningioma grade on external validation through leveraging Cycle-Consistent Adversarial Networks (CycleGAN). In this retrospective study, 257 patients with meningioma were included in the institutional training set. Radiomic features (n = 214) were extracted from T2-weighted (T2) and contrast-enhanced T1 (T1C) images. After radiomics feature selection, extreme gradient boosting classifiers were developed. The models were validated in the external validation set consisting of 61 patients with meningiomas. To reduce the gap in generalization associated with the inter-institutional heterogeneity of MRI, the smaller image set style of the external validation was translated into the larger image set style of the institutional training set using CycleGAN. On external validation before CycleGAN application, the performance of the combined T2 and T1C models showed an area under the curve (AUC), accuracy, and F1 score of 0.77 (95% confidence interval 0.63–0.91), 70.7%, and 0.54, respectively. After applying CycleGAN, the performance of the combined T2 and T1C models increased, with an AUC, accuracy, and F1 score of 0.83 (95% confidence interval 0.70–0.97), 73.2%, and 0.59, respectively. Quantitative metrics (by Fréchet Inception Distance) showed that CycleGAN can decrease inter-institutional image heterogeneity while preserving predictive information. In conclusion, leveraging CycleGAN may be helpful to increase the generalizability of a radiomics model in differentiating meningioma grade on external validation.

Deep Neural Networks and Machine Learning Radiomics Modelling for Prediction of Relapse in Mantle Cell Lymphoma

Simple Summary

Mantle cell lymphoma (MCL) is an aggressive lymphoid tumour with a poor prognosis. There exist no routine biomarkers for the early prediction of relapse. Our study compared the potential of radiomics-based machine learning and 3D deep learning models as non-invasive biomarkers to risk-stratify MCL patients, thus promoting precision imaging in clinical oncology.

Abstract

Mantle cell lymphoma (MCL) is a rare lymphoid malignancy with a poor prognosis characterised by frequent relapse and short durations of treatment response. Most patients present with aggressive disease, but there exist indolent subtypes without the need for immediate intervention. The very heterogeneous behaviour of MCL is genetically characterised by the translocation t(11;14)(q13;q32), leading to Cyclin D1 overexpression with distinct clinical and biological characteristics and outcomes. There is still an unfulfilled need for precise MCL prognostication in real-time. Machine learning and deep learning neural networks are rapidly advancing technologies with promising results in numerous fields of application. This study develops and compares the performance of deep learning (DL) algorithms and radiomics-based machine learning (ML) models to predict MCL relapse on baseline CT scans. Five classification algorithms were used, including three deep learning models (3D SEResNet50, 3D DenseNet, and an optimised 3D CNN) and two machine learning models based on K-nearest Neighbor (KNN) and Random Forest (RF). The best performing method, our optimised 3D CNN, predicted MCL relapse with a 70% accuracy, better than the 3D SEResNet50 (62%) and the 3D DenseNet (59%). The second-best performing method was the KNN-based machine learning model (64%) after principal component analysis for improved accuracy. Our optimised CNN developed by ourselves correctly predicted MCL relapse in 70% of the patients on baseline CT imaging. Once prospectively tested in clinical trials with a larger sample size, our proposed 3D deep learning model could facilitate clinical management by precision imaging in MCL.