Robustness of radiomic features in magnetic resonance imaging: review and a phantom study

Beyond Macrostructure: Is There a Role for Radiomics Analysis in Neuroimaging ?

The most commonly used neuroimaging biomarkers of brain structure, particularly in neurodegenerative diseases, have traditionally been summary measurements from ROIs derived from structural MRI, such as volume and thickness. Advances in MR acquisition techniques, including high-field imaging, and emergence of learning-based methods have opened up opportunities to interrogate brain structure in finer detail, allowing investigators to move beyond macrostructural measurements. On the one hand, superior signal contrast has the potential to make appearance-based metrics that directly analyze intensity patterns, such as texture analysis and radiomics features, more reliable. Quantitative MRI, particularly at high-field, can also provide a richer set of measures with greater interpretability. On the other hand, use of neural networks-based techniques has the potential to exploit subtle patterns in images that can now be mined with advanced imaging. Finally, there are opportunities for integration of multimodal data at different spatial scales that is enabled by developments in many of the above techniques-for example, by combining digital histopathology with high-resolution ex-vivo and in-vivo MRI. Some of these approaches are at early stages of development and present their own set of challenges. Nonetheless, they hold promise to drive the next generation of validation and biomarker studies. This article will survey recent developments in this area, with a particular focus on Alzheimer's disease and related disorders. However, most of the discussion is equally relevant to imaging of other neurological disorders, and even to other organ systems of interest. It is not meant to be an exhaustive review of the available literature, but rather presented as a summary of recent trends through the discussion of a collection of representative studies with an eye towards what the future may hold.

Phantom study investigating the repeatability of radiomic features with alteration of image acquisition parameters in magnetic resonance imaging

BACKGROUND

Magnetic resonance imaging (MRI) has many different alterable parameters that affect how an image appears. This is relevant in radiomics which produces quantitative features through analysis of medical images. One significant acknowledged limitation of radiomics is repeatability. This phantom study aims to further investigate the repeatability of radiomic features (RaF), within MRI, across a range of different echo (TE) and repetition times (TR).

METHODS

A phantom was scanned 10 times under identical conditions on a 3T scanner using head coil over 4 months. The TE ranged from 80 to 110 ms while the TR from 3000 to 5000 ms. Radiomics analysis was performed on the same segmented section of the phantom across all TE and TR combinations. Intraclass Correlation Coefficient (ICC) was calculated across the different TE and TR ranges to investigate the repeatability of RaF.

RESULTS

Of 1596 features calculated, 187 features had ICC >0.9 across the range of TE, while 82 features had an ICC >0.9 across a range of TR. 664 had ICC >0.75 across the range of TEs, with 541 across the range of TR values. There was an overlap of 51 features with ICC >0.9.

CONCLUSION

Repeatability of RaF in MRI is dependent on imaging parameters and careful consideration of these, in combination with variable selection, is required when applying radiomics to MRI.

Radiomics from multisite MRI and clinical data to predict clinically significant prostate cancer

BACKGROUND

Magnetic resonance imaging (MRI) is useful in the diagnosis of clinically significant prostate cancer (csPCa). MRI-derived radiomics may support the diagnosis of csPCa.

PURPOSE

To investigate whether adding radiomics from biparametric MRI to predictive models based on clinical and MRI parameters improves the prediction of csPCa in a multisite-multivendor setting.

MATERIAL AND METHODS

Clinical information (PSA, PSA density, prostate volume, and age), MRI reviews (PI-RADS 2.1), and radiomics (histogram and texture features) were retrieved from prospectively included patients examined at different radiology departments and with different MRI systems, followed by MRI-ultrasound fusion guided biopsies of lesions PI-RADS 3-5. Predictive logistic regression models of csPCa (Gleason score ≥7) for the peripheral (PZ) and transition zone (TZ), including clinical data and PI-RADS only, and combined with radiomics, were built and compared using receiver operating characteristic (ROC) curves.

RESULTS

In total, 456 lesions in 350 patients were analyzed. In PZ and TZ, PI-RADS 4-5 and PSA density, and age in PZ, were independent predictors of csPCa in models without radiomics. In models including radiomics, PI-RADS 4-5, PSA density, age, and ADC energy were independent predictors in PZ, and PI-RADS 5, PSA density and ADC mean in TZ. Comparison of areas under the ROC curve (AUC) for the models without radiomics (PZ: AUC = 0.82, TZ: AUC = 0.80) versus with radiomics (PZ: AUC = 0.82, TZ: AUC = 0.82) showed no significant differences (PZ: P = 0.366; TZ: P = 0.171).

CONCLUSION

PSA density and PI-RADS are potent predictors of csPCa. Radiomics do not add significant information to our multisite-multivendor dataset.

Precision Medicine for Nasopharyngeal Cancer-A Review of Current Prognostic Strategies

Nasopharyngeal carcinoma (NPC) is an Epstein-Barr virus (EBV) driven malignancy arising from the nasopharyngeal epithelium. Current treatment strategies depend on the clinical stage of the disease, including the extent of the primary tumour, the extent of nodal disease, and the presence of distant metastasis. With the close association of EBV infection with NPC development, EBV biomarkers have shown promise in predicting treatment outcomes. Among the omic technologies, RNA and miRNA signatures have been widely studied, showing promising results in the research setting to predict treatment response. The transformation of radiology images into measurable features has facilitated the use of radiomics to generate predictive models for better prognostication and treatment selection. Nonetheless, much of this work remains in the research realm, and challenges remain in clinical implementation.

Impacts of Adaptive Statistical Iterative Reconstruction-V and Deep Learning Image Reconstruction Algorithms on Robustness of CT Radiomics Features: Opportunity for Minimizing Radiomics Variability Among Scans of Different Dose Levels

This study aims to investigate the influence of adaptive statistical iterative reconstruction-V (ASIR-V) and deep learning image reconstruction (DLIR) on CT radiomics feature robustness. A standardized phantom was scanned under single-energy CT (SECT) and dual-energy CT (DECT) modes at standard and low (20 and 10 mGy) dose levels. Images of SECT 120 kVp and corresponding DECT 120 kVp-like virtual monochromatic images were generated with filtered back-projection (FBP), ASIR-V at 40% (AV-40) and 100% (AV-100) blending levels, and DLIR algorithm at low (DLIR-L), medium (DLIR-M), and high (DLIR-H) strength levels. Ninety-four features were extracted via Pyradiomics. Reproducibility of features was calculated between standard and low dose levels, between reconstruction algorithms in reference to FBP images, and within scan mode, using intraclass correlation coefficient (ICC) and concordance correlation coefficient (CCC). The average percentage of features with ICC > 0.90 and CCC > 0.90 between the two dose levels was 21.28% and 20.75% in AV-40 images, and 39.90% and 35.11% in AV-100 images, respectively, and increased from 15.43 to 45.22% and from 15.43 to 44.15% with an increasing strength level of DLIR. The average percentage of features with ICC > 0.90 and CCC > 0.90 in reference to FBP images was 26.07% and 25.80% in AV-40 images, and 18.88% and 18.62% in AV-100 images, respectively, and decreased from 27.93 to 17.82% and from 27.66 to 17.29% with an increasing strength level of DLIR. DLIR and ASIR-V algorithms showed low reproducibility in reference to FBP images, while the high-strength DLIR algorithm provides an opportunity for minimizing radiomics variability due to dose reduction.

Free automatic software for quality assurance of computed tomography calibration, edges and radiomics metrics reproducibility

PURPOSE

To develop a QA procedure, easy to use, reproducible and based on open-source code, to automatically evaluate the stability of different metrics extracted from CT images: Hounsfield Unit (HU) calibration, edge characterization metrics (contrast and drop range) and radiomic features.

METHODS

The QA protocol was based on electron density phantom imaging. Home-made open-source Python code was developed for the automatic computation of the metrics and their reproducibility analysis. The impact on reproducibility was evaluated for different radiation therapy protocols, and phantom positions within the field of view and systems, in terms of variability (Shapiro-Wilk test for 15 repeated measurements carried out over three days) and comparability (Bland-Altman analysis and Wilcoxon Rank Sum Test or Kendall Rank Correlation Coefficient).

RESULTS

Regarding intrinsic variability, most metrics followed a normal distribution (88% of HU, 63% of edge parameters and 82% of radiomic features). Regarding comparability, HU and contrast were comparable in all conditions, and drop range only in the same CT scanner and phantom position. The percentages of comparable radiomic features independent of protocol, position and system were 59%, 78% and 54%, respectively. The non-significantly differences in HU calibration curves obtained for two different institutions (7%) translated in comparable Gamma Index G (1 mm, 1%, >99%).

CONCLUSIONS

An automated software to assess the reproducibility of different CT metrics was successfully created and validated. A QA routine proposal is suggested.

ROOD-MRI: Benchmarking the robustness of deep learning segmentation models to out-of-distribution and corrupted data in MRI

Deep artificial neural networks (DNNs) have moved to the forefront of medical image analysis due to their success in classification, segmentation, and detection challenges. A principal challenge in large-scale deployment of DNNs in neuroimage analysis is the potential for shifts in signal-to-noise ratio, contrast, resolution, and presence of artifacts from site to site due to variances in scanners and acquisition protocols. DNNs are famously susceptible to these distribution shifts in computer vision. Currently, there are no benchmarking platforms or frameworks to assess the robustness of new and existing models to specific distribution shifts in MRI, and accessible multi-site benchmarking datasets are still scarce or task-specific. To address these limitations, we propose ROOD-MRI: a novel platform for benchmarking the Robustness of DNNs to Out-Of-Distribution (OOD) data, corruptions, and artifacts in MRI. This flexible platform provides modules for generating benchmarking datasets using transforms that model distribution shifts in MRI, implementations of newly derived benchmarking metrics for image segmentation, and examples for using the methodology with new models and tasks. We apply our methodology to hippocampus, ventricle, and white matter hyperintensity segmentation in several large studies, providing the hippocampus dataset as a publicly available benchmark. By evaluating modern DNNs on these datasets, we demonstrate that they are highly susceptible to distribution shifts and corruptions in MRI. We show that while data augmentation strategies can substantially improve robustness to OOD data for anatomical segmentation tasks, modern DNNs using augmentation still lack robustness in more challenging lesion-based segmentation tasks. We finally benchmark U-Nets and vision transformers, finding robustness susceptibility to particular classes of transforms across architectures. The presented open-source platform enables generating new benchmarking datasets and comparing across models to study model design that results in improved robustness to OOD data and corruptions in MRI.

Midbrain and pons MRI shape analysis and its clinical and CSF correlates in degenerative parkinsonisms: a pilot study

OBJECTIVES

To conduct brainstem MRI shape analysis across neurodegenerative parkinsonisms and control subjects (CS), along with its association with clinical and cerebrospinal fluid (CSF) correlates.

METHODOLOGY

We collected demographic and clinical variables, performed planimetric and shape MRI analyses, and determined CSF neurofilament-light chain (NfL) levels in 84 participants: 11 CS, 12 with Parkinson's disease (PD), 26 with multiple system atrophy (MSA), 21 with progressive supranuclear palsy (PSP), and 14 with corticobasal degeneration (CBD).

RESULTS

MSA featured the most extensive and significant brainstem shape narrowing (that is, atrophy), mostly in the pons. CBD presented local atrophy in several small areas in the pons and midbrain compared to PD and CS. PSP presented local atrophy in small areas in the posterior and upper midbrain as well as the rostral pons compared to MSA. Our findings of planimetric MRI measurements and CSF NfL levels replicated those from previous literature. Brainstem shape atrophy correlated with worse motor state in all parkinsonisms and with higher NfL levels in MSA, PSP, and PD.

CONCLUSION

Atypical parkinsonisms present different brainstem shape patterns which correlate with clinical severity and neuronal degeneration. In MSA, shape analysis could be further explored as a potential diagnostic biomarker. By contrast, shape analysis appears to have a rather limited discriminant value in PSP.

KEY POINTS

• Atypical parkinsonisms present different brainstem shape patterns. • Shape patterns correlate with clinical severity and neuronal degeneration. • In MSA, shape analysis could be further explored as a potential diagnostic biomarker.

Characterization of Breast Tumors from MR Images Using Radiomics and Machine Learning Approaches

Determining histological subtypes, such as invasive ductal and invasive lobular carcinomas (IDCs and ILCs) and immunohistochemical markers, such as estrogen response (ER), progesterone response (PR), and the HER2 protein status is important in planning breast cancer treatment. MRI-based radiomic analysis is emerging as a non-invasive substitute for biopsy to determine these signatures. We explore the effectiveness of radiomics-based and CNN (convolutional neural network)-based classification models to this end. T1-weighted dynamic contrast-enhanced, contrast-subtracted T1, and T2-weighted MR images of 429 breast cancer tumors from 323 patients are used. Various combinations of input data and classification schemes are applied for ER⁺ vs. ER^-, PR⁺ vs. PR^-, HER2⁺ vs. HER2^-, and IDC vs. ILC classification tasks. The best results were obtained for the ER⁺ vs. ER^- and IDC vs. ILC classification tasks, with their respective AUCs reaching 0.78 and 0.73 on test data. The results with multi-contrast input data were generally better than the mono-contrast alone. The radiomics and CNN-based approaches generally exhibited comparable results. ER and IDC/ILC classification results were promising. PR and HER2 classifications need further investigation through a larger dataset. Better results by using multi-contrast data might indicate that multi-parametric quantitative MRI could be used to achieve more reliable classifiers.

The Role of an MRI-Based Radiomic Signature in Predicting Malignancy of Parotid Gland Tumors

The aim of this study was to assess the ability of MRI radiomic features to differentiate between benign parotid gland tumors (BPGT) and malignant parotid gland tumors (MPGT). This retrospective study included 93 patients who underwent MRI examinations of the head and neck region (78 patients presenting unique PGT, while 15 patients presented double PGT). A total of 108 PGT with histological confirmation were eligible for the radiomic analysis and were assigned to a training group (n = 83; 58 BPGT; 25 MPGT) and a testing group (n = 25; 16 BPGT; 9 MPGT). The radiomic features were extracted from 3D segmentations of the PGT on the T2-weighted and fat-saturated, contrast-enhanced T1-weighted images. Following feature reduction techniques, including LASSO regression analysis, a radiomic signature (RS) was built with five radiomic features. The RS presented a good diagnostic performance in differentiating between PGT, achieving an area under the curve (AUC) of 0.852 (p < 0.001) in the training set and 0.786 (p = 0.017) in the testing set. In both datasets, the RS proved to have lower values in the BPGT group as compared to MPGT group (p < 0.001 and p = 0.023, respectively). The multivariate analysis revealed that RS was independently associated with PGT malignancy, together with the ill-defined margin pattern (p = 0.031, p = 0.001, respectively). The complex model, using clinical data, MRI features and the RS, presented a higher diagnostic performance (AUC of 0.976) in comparison to the RS alone. MRI-based radiomic features could be considered potential additional imaging biomarkers able to discriminate between benign and malignant parotid gland tumors.

Robustness of radiomics features of virtual unenhanced and virtual monoenergetic images in dual-energy CT among different imaging platforms and potential role of CT number variability

OBJECTIVES

To evaluate robustness of dual-energy CT (DECT) radiomics features of virtual unenhanced (VUE) image and virtual monoenergetic image (VMI) among different imaging platforms.

METHODS

A phantom with sixteen clinical-relevant densities was scanned on ten DECT platforms with comparable scan parameters. Ninety-four radiomic features were extracted via Pyradiomics from VUE images and VMIs at energy level of 70 keV (VMI_70keV). Test-retest repeatability was assessed by Bland-Altman analysis. Inter-platform reproducibility of VUE images and VMI_70keV was evaluated by coefficient of variation (CV) and quartile coefficient of dispersion (QCD) among platforms, and by intraclass correlation coefficient (ICC) and concordance correlation coefficient (CCC) between platform pairs. The correlation between variability of CT number radiomics reproducibility was estimated.

RESULTS

92.02% and 92.87% of features were repeatable between scan-rescans for VUE images and VMI_70keV, respectively. Among platforms, 11.30% and 28.39% features of VUE images, and 15.16% and 28.99% features of VMI_70keV were with CV < 10% and QCD < 10%. The average percentages of radiomics features with ICC > 0.90 and CCC > 0.90 between platform pairs were 10.00% and 9.86% in VUE images and 11.23% and 11.23% in VMI_70keV. The CT number inter-platform reproducibility using CV and QCD showed negative correlations with percentage of the first-order radiomics features with CV < 10% and QCD < 10%, in both VUE images and VMI_70keV (r² 0.3870-0.6178, all p < 0.001).

CONCLUSIONS

The majority of DECT radiomics features were non-reproducible. The differences in CT number were considered as an indicator of inter-platform DECT radiomics variation. Critical relevance statement: The majority of radiomics features extracted from the VUE images and the VMI70keV were non-reproducible among platforms, while synchronizing energy levels of VMI to reduce the CT number value variability may be a potential way to mitigate radiomics instability.

Tumor-to-bone distance and radiomic features on MRI distinguish intramuscular lipomas from well-differentiated liposarcomas

Background

To develop a machine learning model based on tumor-to-bone distance and radiomic features derived from preoperative MRI images to distinguish intramuscular (IM) lipomas and atypical lipomatous tumors/well-differentiated liposarcomas (ALTs/WDLSs) and compared with radiologists.

Methods

The study included patients with IM lipomas and ALTs/WDLSs diagnosed between 2010 and 2022, and with MRI scans (sequence/field strength: T1-weighted (T1W) imaging at 1.5 or 3.0 Tesla MRI). Manual segmentation of tumors based on the three-dimensional T1W images was performed by two observers to appraise the intra- and interobserver variability. After radiomic features and tumor-to-bone distance were extracted, it was used to train a machine learning model to distinguish IM lipomas and ALTs/WDLSs. Both feature selection and classification steps were performed using Least Absolute Shrinkage and Selection Operator logistic regression. The performance of the classification model was assessed using a tenfold cross-validation strategy and subsequently evaluated using the receiver operating characteristic curve (ROC) analysis. The classification agreement of two experienced musculoskeletal (MSK) radiologists was assessed using the kappa statistics. The diagnosis accuracy of each radiologist was evaluated using the final pathological results as the gold standard. Additionally, we compared the performance of the model and two radiologists in terms of the area under the receiver operator characteristic curves (AUCs) using the Delong’s test.

Results

There were 68 tumors (38 IM lipomas and 30 ALTs/WDLSs). The AUC of the machine learning model was 0.88 [95% CI 0.72–1] (sensitivity, 91.6%; specificity, 85.7%; and accuracy, 89.0%). For Radiologist 1, the AUC was 0.94 [95% CI 0.87–1] (sensitivity, 97.4%; specificity, 90.9%; and accuracy, 95.0%), and as to Radiologist 2, the AUC was 0.91 [95% CI 0.83–0.99] (sensitivity, 100%; specificity, 81.8%; and accuracy, 93.3%). The classification agreement of the radiologists was 0.89 of kappa value (95% CI 0.76–1). Although the AUC of the model was lower than of two experienced MSK radiologists, there was no statistically significant difference between the model and two radiologists (all P > 0.05).

Conclusions

The novel machine learning model based on tumor-to-bone distance and radiomic features is a noninvasive procedure that has the potential for distinguishing IM lipomas from ALTs/WDLSs. The predictive features that suggested malignancy were size, shape, depth, texture, histogram, and tumor-to-bone distance.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13018-023-03718-4.

Multi-omics to predict acute radiation esophagitis in patients with lung cancer treated with intensity-modulated radiation therapy

PURPOSE

The study aimed to predict acute radiation esophagitis (ARE) with grade  ≥ 2 for patients with locally advanced lung cancer (LALC) treated with intensity-modulated radiation therapy (IMRT) using multi-omics features, including radiomics and dosiomics.

METHODS

161 patients with stage IIIA-IIIB LALC who received chemoradiotherapy (CRT) or radiotherapy by IMRT with a prescribed dose from 45 to 70 Gy from 2015 to 2019 were enrolled retrospectively. All the toxicity gradings were given following the Common Terminology Criteria for Adverse Events V4.0. Multi-omics features, including radiomics, dosiomics (including dose-volume histogram dosimetric parameters), were extracted based on the planning CT image and three-dimensional dose distribution. All data were randomly divided into training cohorts (N = 107) and testing cohorts (N = 54). In the training cohorts, features with reliably high outcome relevance and low redundancy were selected under random patient subsampling. Four classification models (using clinical factors (CF) only, using radiomics features (RFs) only, dosiomics features (DFs) only, and the hybrid features (HFs) containing clinical factors, radiomics and dosiomics) were constructed employing the Ridge classifier using two-thirds of randomly selected patients as the training cohort. The remaining patient was treated as the testing cohort. A series of models were built with 30 times training-testing splits. Their performances were assessed using the area under the ROC curve (AUC) and accuracy.

RESULTS

Among all patients, 51 developed ARE grade  ≥ 2, with an incidence of 31.7%. Next, 8990 radiomics and 213 dosiomics features were extracted, and 3, 6, 12, and 13 features remained after feature selection in the CF, DF, RF and DF models, respectively. The RF and HF models achieved similar classification performance, with the training and testing AUCs of 0.796 ± 0.023 (95% confidence interval (CI [0.79, 0.80])/0.744 ± 0.044 (95% CI [0.73, 0.76]) and 0.801 ± 0.022 (95% CI [0.79, 0.81]) (p = 0.74), respectively. The model performances using CF and DF features were poorer, with training and testing AUCs of 0.573 ± 0.026 (95% CI [0.56, 0.58])/ 0.509 ± 0.072 (95% CI [0.48, 0.53]) and 0.679 ± 0.027 (95% CI [0.67, 0.69])/0.604 ± 0.041 (95% CI [0.53, 0.63]) compared with the above two models (p < 0.001), respectively.

CONCLUSIONS

In LALC patients treated with CRT IMRT, the ARE grade  ≥ 2 can be predicted using the pretreatment radiotherapy image features. To predict ARE, the multi-omics features had similar predictability with radiomics features; however, the dosiomics features and clinical factors had a limited classification performance.

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.

Radiomic Features Associated with Lymphoma Development in the Parotid Glands of Patients with Primary Sjögren's Syndrome

Non-Hodgkin Lymphoma (NHL) represents a severe complication and the main cause of morbidity in patients with primary Sjögren's syndrome (pSS). This study aimed to assess the role of textural analysis (TA) in revealing lymphoma-associated imaging parameters in the parotid gland (PG) parenchyma of patients with pSS. This retrospective study included a total of 36 patients (54.93 ± 13.34 years old; 91.6% females) diagnosed with pSS according to the American College of Rheumatology and the European League Against Rheumatism criteria (24 subjects with pSS and no lymphomatous proliferation; 12 subjects with pSS and NHL development in the PG, confirmed by the histopathological analysis). All subjects underwent MR scanning between January 2018 and October 2022. The coronal STIR PROPELLER sequence was employed to segment PG and perform TA using the MaZda5 software. A total of 65 PGs underwent segmentation and texture feature extraction (48 PGs were included in the pSS control group, and 17 PGs were included in the pSS NHL group). Following parameter reduction techniques, univariate analysis, multivariate regression, and receiver operating characteristics (ROC) analysis, the following TA parameters proved to be independently associated with NHL development in pSS: CH4S6_Sum_Variance and CV4S6_Inverse_Difference_Moment, with an area under ROC of 0.800 and 0.875, respectively. The radiomic model (resulting by combining the two previously independent TA features), presented 94.12% sensitivity and 85.42% specificity in differentiating between the two studied groups, reaching the highest area under ROC of 0.931 for the chosen cutoff value of 1.556. This study suggests the potential role of radiomics in revealing new imaging biomarkers that might serve as useful predictors for lymphoma development in patients with pSS. Further research on multicentric cohorts is warranted to confirm the obtained results and the added benefit of TA in risk stratification for patients with pSS.

Deep learning image reconstruction algorithm reduces image noise while alters radiomics features in dual-energy CT in comparison with conventional iterative reconstruction algorithms: a phantom study

OBJECTIVES

To compare image quality between a deep learning image reconstruction (DLIR) algorithm and conventional iterative reconstruction (IR) algorithms in dual-energy CT (DECT) and to assess the impact of these algorithms on radiomics robustness.

METHODS

A phantom with clinical-relevant densities was imaged on seven DECT scanners with the same voxel size using typical abdominal-pelvis examination protocols. On one DECT scanner, raw data were reconstructed using both conventional IR (adaptive statistical iterative reconstruction-V, ASIR-V) and DLIR. Nine sets of corresponding images were generated on other six DECT scanners using scanner-equipped conventional IR. Regions of interest were delineated through rigid registrations. Image quality was compared. Pyradiomics platform was used for radiomics feature extraction. Test-retest repeatability was assessed by Bland-Altman analysis for repeated scans. Inter-reconstruction algorithm reproducibility between conventional IR and DLIR was tested by intraclass correlation coefficient (ICC) and concordance correlation coefficient (CCC). Inter-scanner reproducibility was evaluated by coefficient of variation (CV) and quartile coefficient of dispersion (QCD). Robust features were identified.

RESULTS

DLIR significantly improved image quality. Ninety-four radiomics features were extracted and nine features were considered as robust. 93.87% features were repeatable between repeated scans. ASIR-V images showed higher reproducibility to other conventional IR than DLIR (ICC mean, 0.603 vs 0.558, p = 0.001; CCC mean, 0.554 vs 0.510, p = 0.004). 7.45% and 26.83% features were reproducible among scanners evaluated by CV and QCD, respectively.

CONCLUSIONS

DLIR improves quality of DECT images but may alter radiomics features compared to conventional IR. Nine robust DECT radiomics features were identified.

KEY POINTS

• DLIR improves DECT image quality in terms of signal-to-noise ratio and contrast-to-noise ratio compared with ASIR-V and showed the highest noise reduction rate and lowest peak frequency shift. • Most of radiomics features are repeatable between repeated DECT scans, while inter-reconstruction algorithm reproducibility between conventional IR and DLIR, and inter-scanner reproducibility, are low. • Although DLIR may alter radiomics features compared to IR algorithms, nine radiomics features survived repeatability and reproducibility analysis among DECT scanners and reconstruction algorithms, which allows further validation and clinical-relevant analysis.

Robustness of Radiomic Features: Two-Dimensional versus Three-Dimensional MRI-Based Feature Reproducibility in Lipomatous Soft-Tissue Tumors

This retrospective study aimed to compare the intra- and inter-observer manual-segmentation variability in the feature reproducibility between two-dimensional (2D) and three-dimensional (3D) magnetic-resonance imaging (MRI)-based radiomic features. The study included patients with lipomatous soft-tissue tumors that were diagnosed with histopathology and underwent MRI scans. Tumor segmentation based on the 2D and 3D MRI images was performed by two observers to assess the intra- and inter-observer variability. In both the 2D and the 3D segmentations, the radiomic features were extracted from the normalized images. Regarding the stability of the features, the intraclass correlation coefficient (ICC) was used to evaluate the intra- and inter-observer segmentation variability. Features with ICC > 0.75 were considered reproducible. The degree of feature robustness was classified as low, moderate, or high. Additionally, we compared the efficacy of 2D and 3D contour-focused segmentation in terms of the effects of the stable feature rate, sensitivity, specificity, and diagnostic accuracy of machine learning on the reproducible features. In total, 93 and 107 features were extracted from the 2D and 3D images, respectively. Only 35 features from the 2D images and 63 features from the 3D images were reproducible. The stable feature rate for the 3D segmentation was more significant than for the 2D segmentation (58.9% vs. 37.6%, p = 0.002). The majority of the features for the 3D segmentation had moderate-to-high robustness, while 40.9% of the features for the 2D segmentation had low robustness. The diagnostic accuracy of the machine-learning model for the 2D segmentation was close to that for the 3D segmentation (88% vs. 90%). In both the 2D and the 3D segmentation, the specificity values were equal to 100%. However, the sensitivity for the 2D segmentation was lower than for the 3D segmentation (75% vs. 83%). For the 2D + 3D radiomic features, the model achieved a diagnostic accuracy of 87% (sensitivity, 100%, and specificity, 80%). Both 2D and 3D MRI-based radiomic features of lipomatous soft-tissue tumors are reproducible. With a higher stable feature rate, 3D contour-focused segmentation should be selected for the feature-extraction process.

MRI Radiomics for Predicting Survival in Patients with Locally Advanced Hypopharyngeal Cancer Treated with Concurrent Chemoradiotherapy

A reliable prognostic stratification of patients with locally advanced hypopharyngeal cancer who had been treated with concurrent chemoradiotherapy (CCRT) is crucial for informing tailored management strategies. The purpose of this retrospective study was to develop robust and objective magnetic resonance imaging (MRI) radiomics-based models for predicting overall survival (OS) and progression-free survival (PFS) in this patient population. The study participants included 198 patients (median age: 52.25 years (interquartile range = 46.88-59.53 years); 95.96% men) who were randomly divided into a training cohort (n = 132) and a testing cohort (n = 66). Radiomic parameters were extracted from post-contrast T1-weighted MR images. Radiomic features for model construction were selected from the training cohort using least absolute shrinkage and selection operator-Cox regression models. Prognostic performances were assessed by calculating the integrated area under the receiver operating characteristic curve (iAUC). The ability of radiomic models to predict OS (iAUC = 0.580, 95% confidence interval (CI): 0.558-0.591) and PFS (iAUC = 0.625, 95% CI = 0.600-0.633) was validated in the testing cohort. The combination of radiomic signatures with traditional clinical parameters outperformed clinical variables alone in the prediction of survival outcomes (observed iAUC increments = 0.279 [95% CI = 0.225-0.334] and 0.293 [95% CI = 0.232-0.351] for OS and PFS, respectively). In summary, MRI radiomics has value for predicting survival outcomes in patients with hypopharyngeal cancer treated with CCRT, especially when combined with clinical prognostic variables.

Performance sensitivity analysis of brain metastasis stereotactic radiosurgery outcome prediction using MRI radiomics

Recent studies have used T1w contrast-enhanced (T1w-CE) magnetic resonance imaging (MRI) radiomic features and machine learning to predict post-stereotactic radiosurgery (SRS) brain metastasis (BM) progression, but have not examined the effects of combining clinical and radiomic features, BM primary cancer, BM volume effects, and using multiple scanner models. To investigate these effects, a dataset of n = 123 BMs from 99 SRS patients with 12 clinical features, 107 pre-treatment T1w-CE radiomic features, and BM progression determined by follow-up MRI was used with a random decision forest model and 250 bootstrapped repetitions. Repeat experiments assessed the relative accuracy across primary cancer sites, BM volume groups, and scanner model pairings. Correction for accuracy imbalances across volume groups was investigated by removing volume-correlated features. We found that using clinical and radiomic features together produced the most accurate model with a bootstrap-corrected area under the receiver operating characteristic curve of 0.77. Accuracy also varied by primary cancer site, BM volume, and scanner model pairings. The effect of BM volume was eliminated by removing features at a volume-correlation coefficient threshold of 0.25. These results show that feature type, primary cancer, volume, and scanner model are all critical factors in the accuracy of radiomics-based prognostic models for BM SRS that must be characterised and controlled for before clinical translation.

Evaluation of radiomics feature stability in abdominal monoenergetic photon counting CT reconstructions

Feature stability and standardization remain challenges that impede the clinical implementation of radiomics. This study investigates the potential of spectral reconstructions from photon-counting computed tomography (PCCT) regarding organ-specific radiomics feature stability. Abdominal portal-venous phase PCCT scans of 10 patients in virtual monoenergetic (VM) (keV 40-120 in steps of 10), polyenergetic, virtual non-contrast (VNC), and iodine maps were acquired. Two 2D and 3D segmentations measuring 1 and 2 cm in diameter of the liver, lung, spleen, psoas muscle, subcutaneous fat, and air were obtained for spectral reconstructions. Radiomics features were extracted with pyradiomics. The calculation of feature-specific intraclass correlation coefficients (ICC) was performed by comparing all segmentation approaches and organs. Feature-wise and organ-wise correlations were evaluated. Segmentation-resegmentation stability was evaluated by concordance correlation coefficient (CCC). Compared to non-VM, VM-reconstruction features tended to be more stable. For VM reconstructions, 3D 2 cm segmentation showed the highest average ICC with 0.63. Based on a criterion of ≥ 3 stable organs and an ICC of ≥ 0.75, 12-mainly non-first-order features-are shown to be stable between the VM reconstructions. In a segmentation-resegmentation analysis in 3D 2 cm, three features were identified as stable based on a CCC of > 0.6 in ≥ 3 organs in ≥ 6 VM reconstructions. Certain radiomics features vary between monoenergetic reconstructions and depend on the ROI size. Feature stability was also shown to differ between different organs. Yet, glcm_JointEntropy, gldm_GrayLevelNonUniformity, and firstorder_Entropy could be identified as features that could be interpreted as energy-independent and segmentation-resegmentation stable in this PCCT collective. PCCT may support radiomics feature standardization and comparability between sites.

Toward the determination of sensitive and reliable whole-lung computed tomography features for robust standard radiomics and delta-radiomics analysis in a nonhuman primate model of coronavirus disease 2019

Purpose

We propose a method to identify sensitive and reliable whole-lung radiomic features from computed tomography (CT) images in a nonhuman primate model of coronavirus disease 2019 (COVID-19). Criteria used for feature selection in this method may improve the performance and robustness of predictive models.

Approach

Fourteen crab-eating macaques were assigned to two experimental groups and exposed to either severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) or a mock inoculum. High-resolution CT scans were acquired before exposure and on several post-exposure days. Lung volumes were segmented using a deep-learning methodology, and radiomic features were extracted from the original image. The reliability of each feature was assessed by the intraclass correlation coefficient (ICC) using the mock-exposed group data. The sensitivity of each feature was assessed using the virus-exposed group data by defining a factor R that estimates the excess of variation above the maximum normal variation computed in the mock-exposed group. R and ICC were used to rank features and identify non-sensitive and unstable features.

Results

Out of 111 radiomic features, 43% had excellent reliability ( ICC > 0.90 ), and 55% had either good ( ICC > 0.75 ) or moderate ( ICC > 0.50 ) reliability. Nineteen features were not sensitive to the radiological manifestations of SARS-CoV-2 exposure. The sensitivity of features showed patterns that suggested a correlation with the radiological manifestations.

Conclusions

Features were quantified and ranked based on their sensitivity and reliability. Features to be excluded to create more robust models were identified. Applicability to similar viral pneumonia studies is also possible.

Radiomics analysis of contrast-enhanced CT scans can distinguish between clear cell and non-clear cell renal cell carcinoma in different imaging protocols

Introduction

This study aimed to construct a radiomics-based machine learning (ML) model for differentiation between non-clear cell and clear cell renal cell carcinomas (ccRCC) that is robust against institutional imaging protocols and scanners.

Materials and methods

Preoperative unenhanced (UN), corticomedullary (CM), and excretory (EX) phase CT scans from 209 patients diagnosed with RCCs were retrospectively collected. After the three-dimensional segmentation, 107 radiomics features (RFs) were extracted from the tumor volumes in each contrast phase. For the ML analysis, the cases were randomly split into training and test sets with a 3:1 ratio. Highly correlated RFs were filtered out based on Pearson’s correlation coefficient (r > 0.95). Intraclass correlation coefficient analysis was used to select RFs with excellent reproducibility (ICC ≥ 0.90). The most predictive RFs were selected by the least absolute shrinkage and selection operator (LASSO). A support vector machine algorithm-based binary classifier (SVC) was constructed to predict tumor types and its performance was evaluated based-on receiver operating characteristic curve (ROC) analysis. The “Kidney Tumor Segmentation 2019” (KiTS19) publicly available dataset was used during external validation of the model. The performance of the SVC was also compared with an expert radiologist’s.

Results

The training set consisted of 121 ccRCCs and 38 non-ccRCCs, while the independent internal test set contained 40 ccRCCs and 13 non-ccRCCs. For external validation, 50 ccRCCs and 23 non-ccRCCs were identified from the KiTS19 dataset with the available UN, CM, and EX phase CTs. After filtering out the highly correlated and poorly reproducible features, the LASSO algorithm selected 10 CM phase RFs that were then used for model construction. During external validation, the SVC achieved an area under the ROC curve (AUC) value, accuracy, sensitivity, and specificity of 0.83, 0.78, 0.80, and 0.74, respectively. UN and/or EX phase RFs did not further increase the model’s performance. Meanwhile, in the same comparison, the expert radiologist achieved similar performance with an AUC of 0.77, an accuracy of 0.79, a sensitivity of 0.84, and a specificity of 0.69.

Conclusion

Radiomics analysis of CM phase CT scans combined with ML can achieve comparable performance with an expert radiologist in differentiating ccRCCs from non-ccRCCs.

Robustness of radiomics to variations in segmentation methods in multimodal brain MRI

Radiomics in neuroimaging uses fully automatic segmentation to delineate the anatomical areas for which radiomic features are computed. However, differences among these segmentation methods affect radiomic features to an unknown extent. A scan-rescan dataset (n = 46) of T1-weighted and diffusion tensor images was used. Subjects were split into a sleep-deprivation and a control group. Scans were segmented using four segmentation methods from which radiomic features were computed. First, we measured segmentation agreement using the Dice-coefficient. Second, robustness and reproducibility of radiomic features were measured using the intraclass correlation coefficient (ICC). Last, difference in predictive power was assessed using the Friedman-test on performance in a radiomics-based sleep deprivation classification application. Segmentation agreement was generally high (interquartile range = 0.77–0.90) and median feature robustness to segmentation method variation was higher (ICC > 0.7) than scan-rescan reproducibility (ICC 0.3–0.8). However, classification performance differed significantly among segmentation methods (p < 0.001) ranging from 77 to 84%. Accuracy was higher for more recent deep learning-based segmentation methods. Despite high agreement among segmentation methods, subtle differences significantly affected radiomic features and their predictive power. Consequently, the effect of differences in segmentation methods should be taken into account when designing and evaluating radiomics-based research methods.

Phantom Study on the Robustness of MR Radiomics Features: Comparing the Applicability of 3D Printed and Biological Phantoms

The objectives of our study were to (a) evaluate the feasibility of using 3D printed phantoms in magnetic resonance imaging (MR) in assessing the robustness and repeatability of radiomic parameters and (b) to compare the results obtained from the 3D printed phantoms to metrics obtained in biological phantoms. To this end, three different 3D phantoms were printed: a Hilbert cube (5 × 5 × 5 cm³) and two cubic quick response (QR) code phantoms (a large phantom (large QR) (5 × 5 × 4 cm³) and a small phantom (small QR) (4 × 4 × 3 cm³)). All 3D printed and biological phantoms (kiwis, tomatoes, and onions) were scanned thrice on clinical 1.5 T and 3 T MR with 1 mm and 2 mm isotropic resolution. Subsequent analyses included analyses of several radiomics indices (RI), their repeatability and reliability were calculated using the coefficient of variation (CV), the relative percentage difference (RPD), and the interclass coefficient (ICC) parameters. Additionally, the readability of QR codes obtained from the MR images was examined with several mobile phones and algorithms. The best repeatability (CV ≤ 10%) is reported for the acquisition protocols with the highest spatial resolution. In general, the repeatability and reliability of RI were better in data obtained at 1.5 T (CV = 1.9) than at 3 T (CV = 2.11). Furthermore, we report good agreements between results obtained for the 3D phantoms and biological phantoms. Finally, analyses of the read-out rate of the QR code revealed better texture analyses for images with a spatial resolution of 1 mm than 2 mm. In conclusion, 3D printing techniques offer a unique solution to create textures for analyzing the reliability of radiomic data from MR scans.

Effect of Gray Value Discretization and Image Filtration on Texture Features of the Pancreas Derived from Magnetic Resonance Imaging at 3T

Radiomics of pancreas magnetic resonance (MR) images is positioned well to play an important role in the management of diseases characterized by diffuse involvement of the pancreas. The effect of image pre-processing configurations on these images has been sparsely investigated. Fifteen individuals with definite chronic pancreatitis (an exemplar diffuse disease of the pancreas) and 15 healthy individuals were included in this age- and sex-matched case-control study. MR images of the pancreas were acquired using a single 3T scanner. A total of 93 first-order and second-order texture features of the pancreas were compared between the study groups, by subjecting MR images of the pancreas to 7 image pre-processing configurations related to gray level discretization and image filtration. The studied parameters of intensity discretization did not vary in terms of their effect on the number of significant first-order texture features. The number of statistically significant first-order texture features varied after filtering (7 with the use of logarithm filter and 3 with the use of Laplacian of Gaussian filter with 5 mm σ). Intensity discretization generally affected the number of significant second-order texture features more markedly than filtering. The use of fixed bin number of 16 yielded 42 significant second-order texture features, fixed bin number of 128–38 features, fixed bin width of 6–24 features, and fixed bin width of 42–26 features. The specific parameters of filtration and intensity discretization had differing effects on radiomics signature of the pancreas. Relative discretization with fixed bin number of 16 and use of logarithm filter hold promise as pre-processing configurations of choice in future radiomics studies in diffuse diseases of the pancreas.

Impact of image quality on radiomics applications

Radiomics features extracted from medical images have been widely reported to be useful in the patient specific outcome modeling for variety of assessment and prediction purposes. Successful application of radiomics features as imaging biomarkers, however, is dependent on the robustness of the approach to the variation in each step of the modeling workflow. Variation in the input image quality is one of the main sources that impacts the reproducibility of radiomics analysis when a model is applied to broader range of medical imaging data. The quality of medical image is generally affected by both the scanner related factors such as image acquisition/reconstruction settings and the patient related factors such as patient motion. This article aimed to review the published literatures in this field that reported the impact of various imaging factors on the radiomics features through the change in image quality. The literatures were categorized by different imaging modalities and also tabulated based on the imaging parameters and the class of radiomics features included in the study. Strategies for image quality standardization were discussed based on the relevant literatures and recommendations for reducing the impact of image quality variation on the radiomics in multi-institutional clinical trial were summarized at the end of this article.

Radiomic Features From Diffusion-Weighted MRI of Retroperitoneal Soft-Tissue Sarcomas Are Repeatable and Exhibit Change After Radiotherapy

Background

Size-based assessments are inaccurate indicators of tumor response in soft-tissue sarcoma (STS), motivating the requirement for new response imaging biomarkers for this rare and heterogeneous disease. In this study, we assess the test-retest repeatability of radiomic features from MR diffusion-weighted imaging (DWI) and derived maps of apparent diffusion coefficient (ADC) in retroperitoneal STS and compare baseline repeatability with changes in radiomic features following radiotherapy (RT).

Materials and Methods

Thirty patients with retroperitoneal STS received an MR examination prior to treatment, of whom 23/30 were investigated in our repeatability analysis having received repeat baseline examinations and 14/30 patients were investigated in our post-treatment analysis having received an MR examination after completing pre-operative RT. One hundred and seven radiomic features were extracted from the full manually delineated tumor region using PyRadiomics. Test-retest repeatability was assessed using an intraclass correlation coefficient (baseline ICC), and post-radiotherapy variance analysis (post-RT-IMS) was used to compare the change in radiomic feature value to baseline repeatability.

Results

For the ADC maps and DWI images, 101 and 102 features demonstrated good baseline repeatability (baseline ICC > 0.85), respectively. Forty-three and 2 features demonstrated both good baseline repeatability and a high post-RT-IMS (>0.85), respectively. Pearson correlation between the baseline ICC and post-RT-IMS was weak (0.432 and 0.133, respectively).

Conclusions

The ADC-based radiomic analysis shows better test-retest repeatability compared with features derived from DWI images in STS, and some of these features are sensitive to post-treatment change. However, good repeatability at baseline does not imply sensitivity to post-treatment change.

Image resampling and discretization effect on the estimate of myocardial radiomic features from T1 and T2 mapping in hypertrophic cardiomyopathy

Radiomics is emerging as a promising and useful tool in cardiac magnetic resonance (CMR) imaging applications. Accordingly, the purpose of this study was to investigate, for the first time, the effect of image resampling/discretization and filtering on radiomic features estimation from quantitative CMR T1 and T2 mapping. Specifically, T1 and T2 maps of 26 patients with hypertrophic cardiomyopathy (HCM) were used to estimate 98 radiomic features for 7 different resampling voxel sizes (at fixed bin width), 9 different bin widths (at fixed resampling voxel size), and 7 different spatial filters (at fixed resampling voxel size/bin width). While we found a remarkable dependence of myocardial radiomic features from T1 and T2 mapping on image filters, many radiomic features showed a limited sensitivity to resampling voxel size/bin width, in terms of intraclass correlation coefficient (> 0.75) and coefficient of variation (< 30%). The estimate of most textural radiomic features showed a linear significant (p < 0.05) correlation with resampling voxel size/bin width. Overall, radiomic features from T2 maps have proven to be less sensitive to image preprocessing than those from T1 maps, especially when varying bin width. Our results might corroborate the potential of radiomics from T1/T2 mapping in HCM and hopefully in other myocardial diseases.

Personalized auto-segmentation for magnetic resonance imaging guided adaptive radiotherapy of prostate cancer

PURPOSE

Fast and accurate delineation of organs on treatment-fraction images is critical in magnetic resonance imaging-guided adaptive radiotherapy (MRIgART). This study proposes a personalized auto-segmentation (AS) framework to assist online delineation of prostate cancer using MRIgART.

METHODS

Image data from 26 patients diagnosed with prostate cancer and treated using hypofractionated MRIgART (5 fractions per patient) were collected retrospectively. Daily pretreatment T2-weighted MRI was performed using a 1.5-T MRI system integrated into a Unity MR-linac. First-fraction image and contour data from 16 patients (80 image-sets) were used to train the population AS model, and the remaining 10 patients composed the test set. The proposed personalized AS framework contained two main steps. First, a convolutional neural network was employed to train the population model using the training set. Second, for each test patient, the population model was progressively fine-tuned with manually checked delineations of the patient's current and previous fractions to obtain a personalized model that was applied to the next fraction.

RESULTS

Compared with the population model, the personalized models substantially improved the mean Dice similarity coefficient from 0.79 to 0.93 for the prostate clinical target volume (CTV), 0.91 to 0.97 for the bladder, 0.82 to 0.92 for the rectum, 0.91 to 0.93 for the femoral heads, respectively.

CONCLUSIONS

The proposed method can achieve accurate segmentation and potentially shorten the overall online delineation time of MRIgART. This article is protected by copyright. All rights reserved.

Automated prediction of the neoadjuvant chemotherapy response in osteosarcoma with deep learning and an MRI-based radiomics nomogram

OBJECTIVES

To implement a pipeline to automatically segment the ROI and to use a nomogram integrating the MRI-based radiomics score and clinical variables to predict responses to neoadjuvant chemotherapy (NAC) in osteosarcoma patients.

METHODS

A total of 144 osteosarcoma patients treated with NAC were separated into training (n = 101) and test (n = 43) groups. After normalisation, ROIs for the preoperative MRI were segmented by a deep learning segmentation model trained with nnU-Net by using two independent manual segmentations as labels. Radiomics features were extracted using automatically segmented ROIs. Feature selection was performed in the training dataset by five-fold cross-validation. The clinical, radiomics, and clinical-radiomics models were built using multiple machine learning methods with the same training dataset and validated with the same test dataset. The segmentation model was evaluated by the Dice coefficient. AUC and decision curve analysis (DCA) were employed to illustrate the model performance and clinical utility.

RESULTS

36/144 (25.0%) patients were pathological good responders (pGRs) to NAC, while 108/144 (75.0%) were non-pGRs. The segmentation model achieved a Dice coefficient of 0.869 on the test dataset. The clinical and radiomics models reached AUCs of 0.636 with a 95% confidence interval (CI) of 0.427-0.860 and 0.759 (95% CI, 0.589-0.937), respectively, in the test dataset. The clinical-radiomics nomogram demonstrated good discrimination, with an AUC of 0.793 (95% CI, 0.610-0.975), and accuracy of 79.1%. The DCA suggested the clinical utility of the nomogram.

CONCLUSION

The automatic nomogram could be applied to aid radiologists in identifying pGRs to NAC.

KEY POINTS

• The nnU-Net trained by manual labels enables the use of an automatic segmentation tool for ROI delineation of osteosarcoma. • A pipeline using automatic lesion segmentation and followed by a radiomics classifier could aid the evaluation of NAC response of osteosarcoma. • A predictive nomogram composed of clinical variables and MRI-based radiomics score provides support for individualised treatment planning.

Challenges in ensuring the generalizability of image quantitation methods for MRI

Image quantitation methods including quantitative MRI, multiparametric MRI, and radiomics offer great promise for clinical use. However, many of these methods have limited clinical adoption, in part due to issues of generalizability, that is, the ability to translate methods and models across institutions. Researchers can assess generalizability through measurement of repeatability and reproducibility, thus quantifying different aspects of measurement variance. In this article, we review the challenges to ensuring repeatability and reproducibility of image quantitation methods as well as present strategies to minimize their variance to enable wider clinical implementation. We present possible solutions for achieving clinically acceptable performance of image quantitation methods and briefly discuss the impact of minimizing variance and achieving generalizability towards clinical implementation and adoption.

Preoperative prediction of lymph node metastasis using deep learning-based features

Lymph node involvement increases the risk of breast cancer recurrence. An accurate non-invasive assessment of nodal involvement is valuable in cancer staging, surgical risk, and cost savings. Radiomics has been proposed to pre-operatively predict sentinel lymph node (SLN) status; however, radiomic models are known to be sensitive to acquisition parameters. The purpose of this study was to develop a prediction model for preoperative prediction of SLN metastasis using deep learning-based (DLB) features and compare its predictive performance to state-of-the-art radiomics. Specifically, this study aimed to compare the generalizability of radiomics vs DLB features in an independent test set with dissimilar resolution. Dynamic contrast-enhancement images from 198 patients (67 positive SLNs) were used in this study. Of these subjects, 163 had an in-plane resolution of 0.7 × 0.7 mm², which were randomly divided into a training set (approximately 67%) and a validation set (approximately 33%). The remaining 35 subjects with a different in-plane resolution (0.78 × 0.78 mm²) were treated as independent testing set for generalizability. Two methods were employed: (1) conventional radiomics (CR), and (2) DLB features which replaced hand-curated features with pre-trained VGG-16 features. The threshold determined using the training set was applied to the independent validation and testing dataset. Same feature reduction, feature selection, model creation procedures were used for both approaches. In the validation set (same resolution as training), the DLB model outperformed the CR model (accuracy 83% vs 80%). Furthermore, in the independent testing set of the dissimilar resolution, the DLB model performed markedly better than the CR model (accuracy 77% vs 71%). The predictive performance of the DLB model outperformed the CR model for this task. More interestingly, these improvements were seen particularly in the independent testing set of dissimilar resolution. This could indicate that DLB features can ultimately result in a more generalizable model.

Precision of MRI radiomics features in the liver and hepatocellular carcinoma

OBJECTIVES

To assess the precision of MRI radiomics features in hepatocellular carcinoma (HCC) tumors and liver parenchyma.

METHODS

The study population consisted of 55 patients, including 16 with untreated HCCs, who underwent two repeat contrast-enhanced abdominal MRI exams within 1 month to evaluate: (1) test-retest repeatability using the same MRI system (n = 28, 10 HCCs); (2) inter-platform reproducibility between different MRI systems (n = 27, 6 HCCs); (3) inter-observer reproducibility (n = 16, 16 HCCs). Shape and 1st- and 2nd-order radiomics features were quantified on pre-contrast T1-weighted imaging (WI), T1WI portal venous phase (pvp), T2WI, and ADC (apparent diffusion coefficient), on liver regions of interest (ROIs) and HCC volumes of interest (VOIs). Precision was assessed by calculating intraclass correlation coefficient (ICC), concordance correlation coefficient (CCC), and coefficient of variation (CV).

RESULTS

There was moderate to excellent test-retest repeatability of shape and 1st- and 2nd-order features for all sequences in HCCs (ICC: 0.53-0.99; CV: 3-29%), and moderate to good test-retest repeatability of 1st- and 2nd-order features for T1WI sequences, and 2nd-order features for T2WI in the liver (ICC: 0.53-0.73; CV: 12-19%). There was poor inter-platform reproducibility for all features and sequences, except for shape and 1st-order features on T1WI in HCCs (CCC: 0.58-0.99; CV: 3-15%). Good to excellent inter-observer reproducibility was found for all features and sequences in HCCs (CCC: 0.80-0.99; CV: 4-15%) and moderate to good for liver (CCC: 0.45-0.86; CV: 6-25%).

CONCLUSIONS

MRI radiomics features have acceptable repeatability in the liver and HCC when using the same MRI system and across readers but have low reproducibility across MR systems, except for shape and 1st-order features on T1WI. Data must be interpreted with caution when performing multiplatform radiomics studies.

KEY POINTS

• MRI radiomics features have acceptable repeatability when using the same MRI system but less reproducible when using different MRI platforms. • MRI radiomics features extracted from T1 weighted-imaging show greater stability across exams than T2 weighted-imaging and ADC. • Inter-observer reproducibility of MRI radiomics features was found to be good in HCC tumors and acceptable in liver parenchyma.

Robustness of CT radiomics features: consistency within and between single-energy CT and dual-energy CT

Objectives

To evaluate inter- and intra- scan mode and scanner repeatability and reproducibility of radiomics features within and between single-energy CT (SECT) and dual-energy CT (DECT).

Methods

A standardized phantom with sixteen rods of clinical-relevant densities was scanned on seven DECT-capable scanners and three SECT-only scanners. The acquisition parameters were selected to present typical abdomen-pelvic examinations with the same voxel size. Images of SECT at 120 kVp and corresponding 120 kVp-like virtual monochromatic images (VMIs) in DECT which were generated according to scanners were analyzed. Regions of interest were drawn with rigid registrations to avoid variations due to segmentation. Radiomics features were extracted via Pyradiomics platform. Test-retest repeatability was evaluated by Bland-Altman analysis for repeated scans. Intra-scanner reproducibility for different scan modes was tested by intraclass correlation coefficient (ICC) and concordance correlation coefficient (CCC). Inter-scanner reproducibility among different scanners for same scan mode was assessed by coefficient of variation (CV) and quartile coefficient of dispersion (QCD).

Results

The test-retest analysis presented that 92.91% and 87.02% of the 94 assessed features were repeatable for SECT 120kVp and DECT 120 kVp-like VMIs, respectively. The intra-scanner analysis for SECT 120kVp vs DECT 120 kVp-like VMIs demonstrated that 10.76% and 10.28% of features were with ICC > 0.90 and CCC > 0.90, respectively. The inter-scanner analysis showed that 17.09% and 27.73% of features for SECT 120kVp were with CV < 10% and QCD < 10%, and 15.16% and 32.78% for DECT 120 kVp-like VMIs, respectively.

Conclusions

The majority of radiomics features were non-reproducible within and between SECT and DECT.

Key Points

• Although the test-retest analysis showed high repeatability for radiomics features, the overall reproducibility of radiomics features within and between SECT and DECT was low.

• Only about one-tenth of radiomics features extracted from SECT images and corresponding DECT images did match each other, even their average photon energy levels were considered alike, indicating that the scan mode potentially altered the radiomics features.

• Less than one-fifth of radiomics features were reproducible among multiple SECT and DECT scanners, regardless of their fixed acquisition and reconstruction parameters, suggesting the necessity of scanning protocol adjustment and post-scan harmonization process.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00330-022-08628-3.

Assessing Robustness of Morphological Characteristics of Arbitrary Grayscale Images

In our previous work, we introduced an empirical model (EM) of arbitrary binary images and three morphological characteristics: disorder of layer structure (DStr), disorder of layer size (DSize), and pattern complexity (PCom). The basic concept of the EM is that forms of lines play no role as a morphological factor in any narrow area of an arbitrary binary image; instead, the basic factor is the type of line connectivity, i.e., isotropic/anisotropic connections. The goal of the present work is to justify the possibility of making the EM applicable for the processing of grayscale arbitrary images. One of the possible ways to reach this goal is to assess the influence of image binarization on the robustness of DStr and DSize. Images that exhibit high and low edge gradient are used for this experimental study. The robustness of DStr and DSize against the binarization procedure is described in absolute (deviation from average) and relative (Pearson’s coefficient correlation) terms. Images with low edge gradient are converted into binary contour maps by applying the watershed algorithm, and DStr and DSize are then calculated for these maps. The robustness of DStr and DSize were assessed against the image threshold for images with high edge gradient and against the grid size of contour maps and Gaussian blur smoothing for images with low edge gradient. Experiments with grayscale arbitrary patterns, such as the surface of Earth and Mars, tidal sand ripples, turbulent flow, a melanoma, and cloud images, are presented to illustrate the spectrum of problems that may be possible to solve by applying the EM. The majority of our experiments show a high level of robustness for DStr and DSize.

Fully automated radiomic screening pipeline for osteoporosis and abnormal bone density with a deep learning-based segmentation using a short lumbar mDixon sequence

BACKGROUND

Although lumbar bone marrow fat fraction (BMFF) has been demonstrated to be predictive of osteoporosis, its utility is limited by the requirement of manual segmentation. Additionally, quantitative features beyond simple BMFF average remain to be explored. In this study, we developed a fully automated radiomic pipeline using deep learning-based segmentation to detect osteoporosis and abnormal bone density (ABD) using a <20 s modified Dixon (mDixon) sequence.

METHODS

In total, 222 subjects underwent quantitative computed tomography (QCT) and lower back magnetic resonance imaging (MRI). Bone mineral density (BMD) were extracted from L1-L3 using QCT as the reference standard; 206 subjects (48.8±14.9 years old, 140 females) were included in the final analysis, and were divided temporally into the training/validation set (142/64 subjects). A deep-learning network was developed to perform automated segmentation. Radiomic models were built using the same training set to predict ABD and osteoporosis using the mDixon maps. The performance was evaluated using the temporal validation set comprised of 64 subjects, along with the automated segmentation. Additional 25 subjects (56.1±8.8 years, 14 females) from another site and a different scanner vendor was included as independent validation to evaluate the performance of the pipeline.

RESULTS

The automated segmentation achieved an outstanding mean dice coefficient of 0.912±0.062 compared to manual in the temporal validation. Task-based evaluation was performed in the temporal validation set, for predicting ABD and osteoporosis, the area under the curve, sensitivity, specificity, and accuracy were 0.925/0.899, 0.923/0.667, 0.789/0.873, 0.844/0.844, respectively. These values were comparable to that of manual segmentation. External validation (cross-vendor) was also performed; the area under the curve, sensitivity, specificity, and accuracy were 0.688/0.913, 0.786/0.857, 0.545/0.944, 0.680/0.920 for ABD and osteoporosis prediction, respectively.

CONCLUSIONS

Our work is the first attempt using radiomics to predict osteoporosis with BMFF map, and the deep-learning based segmentation will further facilitate the clinical utility of the pipeline as a screening tool for early detection of ABD.

Application of Artificial Intelligence in Diagnosis of Craniopharyngioma

Craniopharyngioma is a congenital brain tumor with clinical characteristics of hypothalamic-pituitary dysfunction, increased intracranial pressure, and visual field disorder, among other injuries. Its clinical diagnosis mainly depends on radiological examinations (such as Computed Tomography, Magnetic Resonance Imaging). However, assessing numerous radiological images manually is a challenging task, and the experience of doctors has a great influence on the diagnosis result. The development of artificial intelligence has brought about a great transformation in the clinical diagnosis of craniopharyngioma. This study reviewed the application of artificial intelligence technology in the clinical diagnosis of craniopharyngioma from the aspects of differential classification, prediction of tissue invasion and gene mutation, prognosis prediction, and so on. Based on the reviews, the technical route of intelligent diagnosis based on the traditional machine learning model and deep learning model were further proposed. Additionally, in terms of the limitations and possibilities of the development of artificial intelligence in craniopharyngioma diagnosis, this study discussed the attentions required in future research, including few-shot learning, imbalanced data set, semi-supervised models, and multi-omics fusion.

Foundations of Multiparametric Brain Tumour Imaging Characterisation Using Machine Learning

The heterogeneity of brain tumours at the molecular, metabolic and structural levels poses significant challenge for accurate tissue characterisation. Artificial intelligence and radiomics have emerged as valuable tools to analyse quantitative features extracted from medical images which capture the complex microenvironment of brain tumours. In particular, a number of computational tools including machine learning algorithms have been proposed for image preprocessing, tumour segmentation, feature extraction, classification, and prognostic stratifications as well. In this chapter, we explore the fundamentals of multiparametric brain tumour characterisation, as an understanding of the strengths, limitations and applications of these tools allows clinicians to better develop and evaluate models with improved diagnostic and prognostic value in brain tumour patients.

Robustness of PET Radiomics Features: Impact of Co-Registration with MRI

Radiomics holds great promise in the field of cancer management. However, the clinical application of radiomics has been hampered by uncertainty about the robustness of the features extracted from the images. Previous studies have reported that radiomics features are sensitive to changes in voxel size resampling and interpolation, image perturbation, or slice thickness. This study aims to observe the variability of positron emission tomography (PET) radiomics features under the impact of co-registration with magnetic resonance imaging (MRI) using the difference percentage coefficient, and the Spearman’s correlation coefficient for three groups of images: (i) original PET, (ii) PET after co-registration with T1-weighted MRI and (iii) PET after co-registration with FLAIR MRI. Specifically, seventeen patients with brain cancers undergoing [11C]-Methionine PET were considered. Successively, PET images were co-registered with MRI sequences and 107 features were extracted for each mentioned group of images. The variability analysis revealed that shape features, first-order features and two subgroups of higher-order features possessed a good robustness, unlike the remaining groups of features, which showed large differences in the difference percentage coefficient. Furthermore, using the Spearman’s correlation coefficient, approximately 40% of the selected features differed from the three mentioned groups of images. This is an important consideration for users conducting radiomics studies with image co-registration constraints to avoid errors in cancer diagnosis, prognosis, and clinical outcome prediction.

Evolving Role and Translation of Radiomics and Radiogenomics in Adult and Pediatric Neuro-Oncology

Exponential technologic advancements in imaging, high-performance computing, and artificial intelligence, in addition to increasing access to vast amounts of diverse data, have revolutionized the role of imaging in medicine. Radiomics is defined as a high-throughput feature-extraction method that unlocks microscale quantitative data hidden within standard-of-care medical imaging. Radiogenomics is defined as the linkage between imaging and genomics information. Multiple radiomics and radiogenomics studies performed on conventional and advanced neuro-oncology image modalities show that they have the potential to differentiate pseudoprogression from true progression, classify tumor subgroups, and predict recurrence, survival, and mutation status with high accuracy. In this article, we outline the technical steps involved in radiomics and radiogenomics analyses with the use of artificial intelligence methods and review current applications in adult and pediatric neuro-oncology.

Radiomics feature reliability assessed by intraclass correlation coefficient: a systematic review

Radiomics research is rapidly growing in recent years, but more concerns on radiomics reliability are also raised. This review attempts to update and overview the current status of radiomics reliability research in the ever expanding medical literature from the perspective of a single reliability metric of intraclass correlation coefficient (ICC). To conduct this systematic review, Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed. After literature search and selection, a total of 481 radiomics studies using CT, PET, or MRI, covering a wide range of subject and disease types, were included for review. In these highly heterogeneous studies, feature reliability to image segmentation was much more investigated than reliability to other factors, such as image acquisition, reconstruction, post-processing, and feature quantification. The reported ICCs also suggested high radiomics feature reliability to image segmentation. Image acquisition was found to introduce much more feature variability than image segmentation, in particular for MRI, based on the reported ICC values. Image post-processing and feature quantification yielded different levels of radiomics reliability and might be used to mitigate image acquisition-induced variability. Some common flaws and pitfalls in ICC use were identified, and suggestions on better ICC use were given. Due to the extremely high study heterogeneities and possible risks of bias, the degree of radiomics feature reliability that has been achieved could not yet be safely synthesized or derived in this review. More future researches on radiomics reliability are warranted.

Prospective Evaluation of Repeatability and Robustness of Radiomic Descriptors in Healthy Brain Tissue Regions In Vivo Across Systematic Variations in T2-Weighted Magnetic Resonance Imaging Acquisition Parameters

BACKGROUND

Radiomic descriptors from magnetic resonance imaging (MRI) are promising for disease diagnosis and characterization but may be sensitive to differences in imaging parameters.

OBJECTIVE

To evaluate the repeatability and robustness of radiomic descriptors within healthy brain tissue regions on prospectively acquired MRI scans; in a test-retest setting, under controlled systematic variations of MRI acquisition parameters, and after postprocessing.

STUDY TYPE

Prospective.

SUBJECTS

Fifteen healthy participants.

FIELD STRENGTH/SEQUENCE

A 3.0 T, axial T₂ -weighted 2D turbo spin-echo pulse sequence, 181 scans acquired (2 test/retest reference scans and 12 with systematic variations in contrast weighting, resolution, and acceleration per participant; removing scans with artifacts).

ASSESSMENT

One hundred and forty-six radiomic descriptors were extracted from a contiguous 2D region of white matter in each scan, before and after postprocessing.

STATISTICAL TESTS

Repeatability was assessed in a test/retest setting and between manual and automated annotations for the reference scan. Robustness was evaluated between the reference scan and each group of variant scans (contrast weighting, resolution, and acceleration). Both repeatability and robustness were quantified as the proportion of radiomic descriptors that fell into distinct ranges of the concordance correlation coefficient (CCC): excellent (CCC > 0.85), good (0.7 ≤ CCC ≤ 0.85), moderate (0.5 ≤ CCC < 0.7), and poor (CCC < 0.5); for unprocessed and postprocessed scans separately.

RESULTS

Good to excellent repeatability was observed for 52% of radiomic descriptors between test/retest scans and 48% of descriptors between automated vs. manual annotations, respectively. Contrast weighting (TR/TE) changes were associated with the largest proportion of highly robust radiomic descriptors (21%, after processing). Image resolution changes resulted in the largest proportion of poorly robust radiomic descriptors (97%, before postprocessing). Postprocessing of images with only resolution/acceleration differences resulted in 73% of radiomic descriptors showing poor robustness.

DATA CONCLUSIONS

Many radiomic descriptors appear to be nonrobust across variations in MR contrast weighting, resolution, and acceleration, as well in test-retest settings, depending on feature formulation and postprocessing.

EVIDENCE LEVEL

2 TECHNICAL EFFICACY: Stage 2.

Making Radiomics More Reproducible across Scanner and Imaging Protocol Variations: A Review of Harmonization Methods

Radiomics converts medical images into mineable data via a high-throughput extraction of quantitative features used for clinical decision support. However, these radiomic features are susceptible to variation across scanners, acquisition protocols, and reconstruction settings. Various investigations have assessed the reproducibility and validation of radiomic features across these discrepancies. In this narrative review, we combine systematic keyword searches with prior domain knowledge to discuss various harmonization solutions to make the radiomic features more reproducible across various scanners and protocol settings. Different harmonization solutions are discussed and divided into two main categories: image domain and feature domain. The image domain category comprises methods such as the standardization of image acquisition, post-processing of raw sensor-level image data, data augmentation techniques, and style transfer. The feature domain category consists of methods such as the identification of reproducible features and normalization techniques such as statistical normalization, intensity harmonization, ComBat and its derivatives, and normalization using deep learning. We also reflect upon the importance of deep learning solutions for addressing variability across multi-centric radiomic studies especially using generative adversarial networks (GANs), neural style transfer (NST) techniques, or a combination of both. We cover a broader range of methods especially GANs and NST methods in more detail than previous reviews.

Radiomics of diffusion-weighted MRI compared to conventional measurement of apparent diffusion-coefficient for differentiation between benign and malignant soft tissue tumors

Diffusion-weighted imaging (DWI) is proven useful to differentiate benign and malignant soft tissue tumors (STTs). Radiomics utilizing a vast array of extracted imaging features has a potential to uncover disease characteristics. We aim to assess radiomics using DWI can outperform the conventional DWI for STT differentiation. In 151 patients with 80 benign and 71 malignant tumors, ADC_mean and ADC_min were measured on solid portion within the mass by two different readers. For radiomics approach, tumors were segmented and 100 original radiomic features were extracted on ADC map. Eight radiomics models were built with training set (n = 105), using combinations of 2 different algorithms—multivariate logistic regression (MLR) and random forest (RF)—and 4 different inputs: radiomics features (R), R + ADC_min (I), R + ADC_mean (E), R + ADC_min and ADC_mean (A). All models were validated with test set (n = 46), and AUCs of ADC_mean, ADC_min, MLR-R, RF-R, MLR-I, RF-I, MLR-E, RF-E, MLR-A and RF-A models were 0.729, 0.753 0.698, 0.700, 0.773, 0.807, 0.762, 0.744, 0.773 and 0.807, respectively, without statistically significant difference. In conclusion, radiomics approach did not add diagnostic value to conventional ADC measurement for differentiating benign and malignant STTs.

MRI-derived radiomics model for baseline prediction of prostate cancer progression on active surveillance

Nearly half of patients with prostate cancer (PCa) harbour low- or intermediate-risk disease considered suitable for active surveillance (AS). However, up to 44% of patients discontinue AS within the first five years, highlighting the unmet clinical need for robust baseline risk-stratification tools that enable timely and accurate prediction of tumour progression. In this proof-of-concept study, we sought to investigate the added value of MRI-derived radiomic features to standard-of-care clinical parameters for improving baseline prediction of PCa progression in AS patients. Tumour T₂-weighted imaging (T2WI) and apparent diffusion coefficient radiomic features were extracted, with rigorous calibration and pre-processing methods applied to select the most robust features for predictive modelling. Following leave-one-out cross-validation, the addition of T2WI-derived radiomic features to clinical variables alone improved the area under the ROC curve for predicting progression from 0.61 (95% confidence interval [CI] 0.481-0.743) to 0.75 (95% CI 0.64-0.86). These exploratory findings demonstrate the potential benefit of MRI-derived radiomics to add incremental benefit to clinical data only models in the baseline prediction of PCa progression on AS, paving the way for future multicentre studies validating the proposed model and evaluating its impact on clinical outcomes.

Role of Radiomics in the Prediction of Muscle-invasive Bladder Cancer: A Systematic Review and Meta-analysis

CONTEXT

Radiomics is a field of science that aims to develop improved methods of medical image analysis by extracting a large number of quantitative features. New data have emerged on the successful application of radiomics and machine-learning techniques to the prediction of muscle-invasive bladder cancer (MIBC).

OBJECTIVE

To systematically review the diagnostic performance of radiomic techniques in predicting MIBC.

EVIDENCE ACQUISITION

The literature search for relevant studies up to July 2020 was performed in the PubMed and EMBASE databases by two independent reviewers. The meta-analysis was inducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines. Inclusion criteria comprised studies that evaluated the diagnostic accuracy of radiomic models in predicting MIBC and used pathological examination as the reference standard. For bias assessment, Quality Assessment of Diagnostic Accuracy Studies-2 and Radiomic Quality Score were used. Weighted summary proportions were used to calculate pooled sensitivity and specificity. A linear mixed model was implemented to calculate the hierarchical summary receiver-operating characteristic (HSROC). Meta-regression analyses were performed to explore heterogeneity.

EVIDENCE SYNTHESIS

Eight studies with a total of 860 patients were included. The summary estimates for sensitivity and specificity in predicting MIBC were 82% (95% confidence interval [CI]: 77-86%) and 81% (95% CI: 76-85%), respectively. The area under HSROC was 0.88. There were no relevant heterogeneity in diagnostic accuracy measures (I² = 33% and 41% for sensitivity and specificity, respectively), which was confirmed by a subsequent meta-regression analysis.

CONCLUSIONS

Radiomics shows high diagnostic performance in predicting MIBC. Despite differences in approaches, radiomic models were relatively homogeneous in their diagnostic accuracy. With further improvements, radiomics has the potential to become a useful adjunct in clinical management of bladder cancer.

PATIENT SUMMARY

Rapidly evolving imaging analysis methods using artificial intelligence algorithms, called radiomics, show high diagnostic performance in predicting muscle-invasive bladder cancer.

Challenges and opportunities for artificial intelligence in oncological imaging

Imaging plays a key role in oncology, including the diagnosis and detection of cancer, determining clinical management, assessing treatment response, and complications of treatment or disease. The current use of clinical oncology is predominantly qualitative in nature with some relatively crude size-based measurements of tumours for assessment of disease progression or treatment response; however, it is increasingly understood that there may be significantly more information about oncological disease that can be obtained from imaging that is not currently utilized. Artificial intelligence (AI) has the potential to harness quantitative techniques to improve oncological imaging. These may include improving the efficiency or accuracy of traditional roles of imaging such as diagnosis or detection. These may also include new roles for imaging such as risk-stratifying patients for different types of therapy or determining biological tumour subtypes. This review article outlines several major areas in oncological imaging where there may be opportunities for AI technology. These include (1) screening and detection of cancer, (2) diagnosis and risk stratification, (3) tumour segmentation, (4) precision oncology, and (5) predicting prognosis and assessing treatment response. This review will also address some of the potential barriers to AI research in oncological imaging.

Robustness of radiomic features in CT images with different slice thickness, comparing liver tumour and muscle

Radiomic image features are becoming a promising non-invasive method to obtain quantitative measurements for tumour classification and therapy response assessment in oncological research. However, despite its increasingly established application, there is a need for standardisation criteria and further validation of feature robustness with respect to imaging acquisition parameters. In this paper, the robustness of radiomic features extracted from computed tomography (CT) images is evaluated for liver tumour and muscle, comparing the values of the features in images reconstructed with two different slice thicknesses of 2.0 mm and 5.0 mm. Novel approaches are presented to address the intrinsic dependencies of texture radiomic features, choosing the optimal number of grey levels and correcting for the dependency on volume. With the optimal values and corrections, feature values are compared across thicknesses to identify reproducible features. Normalisation using muscle regions is also described as an alternative approach. With either method, a large fraction of features (75-90%) was found to be highly robust (< 25% difference). The analyses were performed on a homogeneous CT dataset of 43 patients with hepatocellular carcinoma, and consistent results were obtained for both tumour and muscle tissue. Finally, recommended guidelines are included for radiomic studies using variable slice thickness.