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

Correlations of characteristics with tissue involvement in knee gouty arthritis: Magnetic resonance imaging analysis

Objective

This study investigates the MRI features of knee gouty arthritis (KGA), examines its relationship with the extent of tissue involvement, and assesses whether risk factors can predict KGA.

Materials and methods

Patients diagnosed with KGA underwent MRI examinations, and two independent observers retrospectively analyzed data from 44 patients (49 knees). These patients were divided into mild and severe groups based on tissue involvement observed during arthroscopy. MRI features were summarized, and the intraclass correlation coefficient evaluated interobserver reproducibility. Single-factor analysis compared clinical indicators and MRI features between groups, while Cramer's V coefficient assessed correlations. Multivariate logistic regression identified predictors of tissue involvement extent, and a ROC curve evaluated diagnostic performance.

Results

Among 49 knees, 18 had mild and 31 had severe tissue involvement. Key MRI features included ligament sketch-like changes, meniscal urate deposition, irregularly serrated cartilage changes, low-signal signs within joint effusion, synovial proliferation, Hoffa's fat pad synovitis, gouty tophi, bone erosion, and bone marrow edema. The interobserver reliability of the MRI features was good. Significant differences (P < 0.05) were observed between the groups for anterior cruciate ligament (ACL) sketch-like changes, Hoffa's fat pad synovitis, and gouty tophi. ACL sketch-like changes (r = 0.309), Hoffa's fat pad synovitis (r = 0.309), and gouty tophi (r = 0.408) were positively correlated with the extent of tissue involvement (P < 0.05). ACL sketch-like changes (OR = 9.019, 95 % CI: 1.364-61.880), Hoffa's fat pad synovitis (OR = 6.472, 95 % CI: 1.041-40.229), and gouty tophi (OR = 5.972, 95 % CI: 1.218-29.276) were identified as independent predictors of tissue involvement extent (P < 0.05). The area under the ROC curve was 0.862, with a sensitivity of 67.70 %, specificity of 94.40 %, and accuracy of 79.14 %.

Conclusion

This comprehensive analysis of MRI features identifies ligament sketch-like changes, meniscal urate deposition, and low-signal signs within joint effusion as characteristic MRI manifestations of KGA. Irregular cartilage changes are valuable for differential diagnosis in young and middle-aged patients. ACL sketch-like changes, Hoffa's fat pad synovitis, and gouty tophi correlate with tissue involvement severity and are critical in predicting and assessing the extent of tissue involvement in KGA.

Does consensus contour improve robustness and accuracy in 18F-FDG PET radiomic features?

PURPOSE

The purpose of our study is to validate the robustness and accuracy of consensus contour in 2-deoxy-2-[18 F]fluoro-D-glucose (18 F-FDG) PET radiomic features.

METHODS

225 nasopharyngeal carcinoma (NPC) and 13 extended cardio-torso (XCAT) simulated data were enrolled. All segmentation were performed with four segmentation methods under two different initial masks, respectively. Consensus contour (ConSeg) was then developed using the majority vote rule. 107 radiomic features were extracted by Pyradiomics based on segmentation and the intraclass correlation coefficient (ICC) was calculated for each feature between masks or among segmentation, respectively. In XCAT ICC between segmentation and simulated ground truth were also calculated to access the accuracy.

RESULTS

ICC varied with the dataset, segmentation method, initial mask and feature type. ConSeg presented higher ICC for radiomic features in robustness tests and similar ICC in accuracy tests, compared with the average of four segmentation results. Higher ICC were also generally observed in irregular initial masks compared with rectangular masks in both robustness and accuracy tests. Furthermore, 19 features (17.76%) had ICC ≥ 0.75 in both robustness and accuracy tests for any of the segmentation methods or initial masks. The dataset was observed to have a large impact on the correlation relationships between radiomic features, but not the segmentation method or initial mask.

CONCLUSIONS

The consensus contour combined with irregular initial mask could improve the robustness and accuracy in radiomic analysis to some extent. The correlation relationships between radiomic features and feature clusters largely depended on the dataset, but not segmentation method or initial mask.

Establishment and validation of novel predictive models to predict bone metastasis in newly diagnosed prostate adenocarcinoma based on single-photon emission computed tomography radiomics

PURPOSE

To establish and validate novel predictive models for predicting bone metastasis (BM) in newly diagnosed prostate adenocarcinoma (PCa) via single-photon emission computed tomography radiomics.

METHOD

In a retrospective review of the clinical single-photon emission computed tomography (SPECT) database, 176 patients (training set: n = 140; validation set: n = 36) who underwent SPECT/CT imaging and were histologically confirmed to have newly diagnosed PCa from June 2016 to June 2022 were enrolled. Radiomic features were extracted from the region of interest (ROI) in a targeted lesion in each patient. Clinical features, including age, total prostate-specific antigen (t-PSA), and Gleason grades, were included. Statistical tests were then employed to eliminate irrelevant and redundant features. Finally, four types of optimized models were constructed for the prediction. Furthermore, fivefold cross-validation was applied to obtain sensitivity, specificity, accuracy, and area under the curve (AUC) for performance evaluation. The clinical usefulness of the multivariate models was estimated through decision curve analysis (DCA).

RESULTS

A radiomics signature consisting of 27 selected features which were obtained by radiomics' LASSO treatment was significantly correlated with bone status (P < 0.01 for both training and validation sets). Collectively, the models showed good predictive efficiency. The AUC values ranged from 0.87 to 0.98 in four models. The AUC values of the human experts were 0.655 and 0.872 in the training and validation groups, respectively. Most radiomic models showed better diagnostic accuracy than human experts in the training and validation groups. DCA also demonstrated the superiority of the radiomics models compared to human experts.

CONCLUSION

Radiomics models are superior to humans in differentiating between benign bone and prostate cancer bone metastases; it can be used to facilitate personalized prediction of BM in newly diagnosed PCa patients.

Novel AI-based tool for primary tooth segmentation on CBCT using convolutional neural networks: A validation study

BACKGROUND

Primary teeth segmentation on cone beam computed tomography (CBCT) scans is essential for paediatric treatment planning. Conventional methods, however, are time-consuming and necessitate advanced expertise.

AIM

The aim of this study was to validate an artificial intelligence (AI) cloud-based platform for automated segmentation (AS) of primary teeth on CBCT. Its accuracy, time efficiency, and consistency were compared with manual segmentation (MS).

DESIGN

A dataset comprising 402 primary teeth (37 CBCT scans) was retrospectively retrieved from two CBCT devices. Primary teeth were manually segmented using a cloud-based platform representing the ground truth, whereas AS was performed on the same platform. To assess the AI tool's performance, voxel- and surface-based metrics were employed to compare MS and AS methods. Additionally, segmentation time was recorded for each method, and intra-class correlation coefficient (ICC) assessed consistency between them.

RESULTS

AS revealed high performance in segmenting primary teeth with high accuracy (98 ± 1%) and dice similarity coefficient (DSC; 95 ± 2%). Moreover, it was 35 times faster than the manual approach with an average time of 24 s. Both MS and AS demonstrated excellent consistency (ICC = 0.99 and 1, respectively).

CONCLUSION

The platform demonstrated expert-level accuracy, and time-efficient and consistent segmentation of primary teeth on CBCT scans, serving treatment planning in children.

Role of radiomics as a predictor of disease recurrence in ovarian cancer: a systematic review

Ovarian cancer is associated with high cancer-related mortality rate attributed to late-stage diagnosis, limited treatment options, and frequent disease recurrence. As a result, careful patient selection is important especially in setting of radical surgery. Radiomics is an emerging field in medical imaging, which may help provide vital prognostic evaluation and help patient selection for radical treatment strategies. This systematic review aims to assess the role of radiomics as a predictor of disease recurrence in ovarian cancer. A systematic search was conducted in Medline, EMBASE, and Web of Science databases. Studies meeting inclusion criteria investigating the use of radiomics to predict post-operative recurrence in ovarian cancer were included in our qualitative analysis. Study quality was assessed using the QUADAS-2 and Radiomics Quality Score tools. Six retrospective studies met the inclusion criteria, involving a total of 952 participants. Radiomic-based signatures demonstrated consistent performance in predicting disease recurrence, as evidenced by satisfactory area under the receiver operating characteristic curve values (AUC range 0.77-0.89). Radiomic-based signatures appear to good prognosticators of disease recurrence in ovarian cancer as estimated by AUC. The reviewed studies consistently reported the potential of radiomic features to enhance risk stratification and personalise treatment decisions in this complex cohort of patients. Further research is warranted to address limitations related to feature reliability, workflow heterogeneity, and the need for prospective validation studies.

MRI radiomics may predict early tumor recurrence in patients with sinonasal squamous cell carcinoma

OBJECTIVES

Sinonasal squamous cell carcinoma (SCC) follows a poor prognosis with high tendency for local recurrence. We aimed to evaluate whether MRI radiomics can predict early local failure in sinonasal SCC.

METHODS

Sixty-eight consecutive patients with node-negative sinonasal SCC (January 2005-December 2020) were enrolled, allocated to the training (n = 47) and test sets (n = 21). Early local failure, which occurred within 12 months of completion of initial treatment, was the primary endpoint. For clinical features (age, location, treatment modality, and clinical T stage), binary logistic regression analysis was performed. For 186 extracted radiomic features, different feature selections and classifiers were combined to create two prediction models: (1) a pure radiomics model; and (2) a combined model with clinical features and radiomics. The areas under the receiver operating characteristic curves (AUCs) were calculated and compared using DeLong's method.

RESULTS

Early local failure occurred in 38.3% (18/47) and 23.8% (5/21) in the training and test sets, respectively. We identified several radiomic features which were strongly associated with early local failure. In the test set, both the best-performing radiomics model and the combined model (clinical + radiomic features) yielded higher AUCs compared to the clinical model (AUC, 0.838 vs. 0.438, p = 0.020; 0.850 vs. 0.438, p = 0.016, respectively). The performances of the best-performing radiomics model and the combined model did not differ significantly (AUC, 0.838 vs. 0.850, p = 0.904).

CONCLUSION

MRI radiomics integrated with a machine learning classifier may predict early local failure in patients with sinonasal SCC.

CLINICAL RELEVANCE STATEMENT

MRI radiomics intergrated with machine learning classifiers may predict early local failure in sinonasal squamous cell carcinomas more accurately than the clinical model.

KEY POINTS

• A subset of radiomic features which showed significant association with early local failure in patients with sinonasal squamous cell carcinomas was identified. • MRI radiomics integrated with machine learning classifiers can predict early local failure with high accuracy, which was validated in the test set (area under the curve = 0.838). • The combined clinical and radiomics model yielded superior performance for early local failure prediction compared to that of the radiomics (area under the curve 0.850 vs. 0.838 in the test set), without a statistically significant difference.

Radiomic features of the hippocampal based on magnetic resonance imaging in the menopausal mouse model linked to neuronal damage and cognitive deficits

Estrogen deficiency in the early postmenopausal phase is associated with an increased long-term risk of cognitive decline or dementia. Non-invasive characterization of the pathological features of the pathological hallmarks in the brain associated with postmenopausal women (PMW) could enhance patient management and the development of therapeutic strategies. Radiomics is a means to quantify the radiographic phenotype of a diseased tissue via the high-throughput extraction and mining of quantitative features from images acquired from modalities such as CT and magnetic resonance imaging (MRI). This study set out to explore the correlation between radiomics features based on MRI and pathological features of the hippocampus and cognitive function in the PMW mouse model. Ovariectomized (OVX) mice were used as PWM models. MRI scans were performed two months after surgery. The brain's hippocampal region was manually annotated, and the radiomic features were extracted with PyRadiomics. Chemiluminescence was used to evaluate the peripheral blood estrogen level of mice, and the Morris water maze test was used to evaluate the cognitive ability of mice. Nissl staining and immunofluorescence were used to quantify neuronal damage and COX1 expression in brain sections of mice. The OVX mice exhibited marked cognitive decline, brain neuronal damage, and increased expression of mitochondrial complex IV subunit COX1, which are pathological phenomena commonly observed in the brains of AD patients, and these phenotypes were significantly correlated with radiomics features (p < 0.05, |r|>0.5), including Original_firstorder_Interquartile Range, Original_glcm_Difference Average, Original_glcm_Difference Average and Wavelet-LHH_glszm_Small Area Emphasis. Meanwhile, the above radiomics features were significantly different between the sham-operated and OVX groups (p < 0.01) and were associated with decreased serum estrogen levels (p < 0.05, |r|>0.5). This initial study indicates that the above radiomics features may have a role in the assessment of the pathology of brain damage caused by estrogen deficiency using routinely acquired structural MR images.

Multi-sequence MRI radiomics of colorectal liver metastases: Which features are reproducible across readers?

PURPOSE

To assess the inter-reader reproducibility of radiomics features on multiple MRI sequences after segmentations of colorectal liver metastases (CRLM).

METHOD

30 CRLM (in 23 patients) were manually delineated by three readers on MRI before the start of chemotherapy on the contrast enhanced T1-weighted images (CE-T1W) in the portal venous phase, T2-weighted images (T2W) and b800 diffusion weighted images (DWI). DWI delineations were copied to the ADC-maps. 107 radiomics features were extracted per sequence. The intraclass correlation coefficient (ICC) was calculated per feature. Features were considered reproducible if ICC > 0.9.

RESULTS

90% of CE-T1W features were reproducible with a median ICC of 0.98 (range 0.76-1.00). 81% of DWI features were robust with median ICC = 0.97 (range 0.38-1.00). The T2W features had a median ICC of 0.96 (range 0.55-0.99) and were reproducible in 80%. ADC showed the lowest number of reproducible features with 58% and median ICC = 0.91 (range 0.38-0.99) When considering the lower bound of the ICC 95% confidence intervals, 58%, 66%, 54% and 29% reached 0.9 for the CE-T1W, DWI, T2W and ADC features, respectively. The feature class with the best reproducibility differed per sequence.

CONCLUSIONS

The majority of MRI radiomics features from CE-T1W, T2W, DWI and ADC in colorectal liver metastases were robust for segmentation variability between readers. The CE-T1W yielded slightly better reproducibility results compared to DWI and T2W. The ADC features seem more susceptible to reader differences compared to the other three sequences.

Interobserver delineation variability of computed tomography-based radiomic features of the parotid gland

PURPOSE

To assess the interobserver delineation variability of radiomic features of the parotid gland from computed tomography (CT) images and evaluate the correlation of these features for head and neck cancer (HNC) radiotherapy patients.

MATERIALS AND METHODS

Contrast-enhanced CT images of 20 HNC patients were utilized. The parotid glands were delineated by treating radiation oncologists (ROs), a selected RO and AccuContour auto-segmentation software. Dice similarity coefficients (DSCs) between each pair of observers were calculated. A total of 107 radiomic features were extracted, whose robustness to interobserver delineation was assessed using the intraclass correlation coefficient (ICC). Pearson correlation coefficients (r) were calculated to determine the relationship between the features. The influence of excluding unrobust features from normal tissue complication probability (NTCP) modeling was investigated for severe oral mucositis (grade ≥3).

RESULTS

The average DSC was 0.84 (95% confidence interval, 0.83-0.86). Most of the shape features demonstrated robustness (ICC ≥0.75), while the first-order and texture features were influenced by delineation variability. Among the three observers investigated, 42 features were sufficiently robust, out of which 36 features exhibited weak correlation (|r|<0.8). No significant difference in the robustness level was found when comparing manual segmentation by a single RO or automated segmentation with the actual clinical contour data made by treating ROs. Excluding unrobust features from the NTCP model for severe oral mucositis did not deteriorate the model performance.

CONCLUSION

Interobserver delineation variability had substantial impact on radiomic features of the parotid gland. Both manual and automated segmentation methods contributed similarly to this variation.

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.

Radiomics from dual-energy CT-derived iodine maps predict lymph node metastasis in head and neck squamous cell carcinoma

OBJECTIVE

To develop and validate an iodine maps-based radiomics nomogram for preoperatively predicting cervical lymph node metastasis (LNM) in head and neck squamous cell carcinoma (HNSCC).

MATERIALS AND METHODS

A total of 278 patients who pathologically confirmed as HNSCC were retrospectively recruited from two medical centers between June 2012 and July 2022. The training set (n = 152) and internal set (n = 67) were randomly selected from medical center A, and the patients from medical center B were enrolled as the external set (n = 69). The minority group in the training set was balanced by the adaptive synthetic sampling (ADASYN) approach. Radiomics features were extracted from dual-energy CT-derived iodine maps at arterial phase (AP) and venous phase (VP), respectively. Three radiomics signatures were constructed to predict the LNM by using a random forest algorithm. The independent clinical predictors for LNM were identified by multivariate analysis and combined with radiomics signatures to establish a radiomic-clinical nomogram. The performance of radiomic-clinical nomogram was evaluated with respect to its discrimination and clinical usefulness.

RESULTS

The AP-VP-incorporated radiomics model exhibited a great predictive performance for LNM prediction with an area under curve (AUC) of 0.885 (95% CI, 0.836-0.933) in ADASYN-training set and confirmed in all validation sets. The nomogram that incorporated AP-VP radiomics signatures, CT-reported LN status, and histological grades yielded AUCs of 0.920 (95% CI, 0.881-0.959) in ADASYN-training set, 0.858 (95% CI, 0.771-0.944) in internal validation, and 0.849 (95% CI, 0.752-0.946) in external validation, with good calibration in all cohorts (p > 0.05). Decision curve analyses indicated the nomogram was clinically useful. In addition, the predictive performance of clinical-radiomics nomogram was also validation in combing cohorts. Stratified analysis confirmed the stability of nomogram, particularly in group negative for CT-reported LNM.

CONCLUSION

Clinical-radiomics nomogram based on iodine maps exhibited promising performance in predicting LNM and providing valuable information for making individualized therapy decisions.

Deep learning and machine learning predictive models for neurological function after interventional embolization of intracranial aneurysms

Objective

The objective of this study is to develop a model to predicts the postoperative Hunt-Hess grade in patients with intracranial aneurysms by integrating radiomics and deep learning technologies, using preoperative CTA imaging data. Thereby assisting clinical decision-making and improving the assessment and prognosis of postoperative neurological function.

Methods

This retrospective study encompassed 101 patients who underwent aneurysm embolization surgery. 851 radiomic features were extracted from CTA images. 512 deep learning features are extracted from last layer of ResNet50 deep convolutional neural network model. The feature screening process pipeline encompassed intraclass correlation coefficient analysis, principal component analysis, U test, spearman correlation analysis, minimum redundancy maximum relevance algorithm and Lasso regression, to identify features most correlated with postoperative Hunt-Hess grading. In the model construction phase, three distinct models were constructed: radiomics feature-based model (RSM), deep learning feature-based model (DLM), and deep learning-radiomics feature fusion model (DLRSCM). The study also calculated the radiomics score and combined it with clinical data to construct a Nomogram for predictive modeling. DLM, RSM and DLRSCM model was constructed by 9 base algorithms and 1 ensemble learning algorithm - Stacking ensemble model. Model performance was evaluated based on the area under the Receiver Operating Characteristic (ROC) curve (AUC), Matthews Correlation Coefficient (MCC), calibration curves, and decision curves analysis.

Results

5 significant radiomic feature and 4 significant deep learning features were obtained through the feature selection process. These features were utilized for model construction. Bootstrap resampling method was used for internal validation of the models. In terms of model evaluation, the DLM model, the stacking ensemble algorithm results achieved an AUC of 0.959 and MCC of 0.815. In the RSM model, the stacking ensemble model AUC was 0.935 and MCC was 0.793. The stacking ensemble model in DLRSCM outperformed others, with an AUC of 0.968 and MCC of 0.820. Results indicated that the ANN performed optimally among all base models, while the stacked ensemble learning model exhibited the highest predictive performance.

Conclusion

This study demonstrates that the combination of radiomics and deep learning is an effective approach to predict the postoperative Hunt-Hess grade in patients with intracranial aneurysms. This holds significant value in the early identification of postoperative neurological complications and in enhancing clinical decision-making.

MRI-based interpretable radiomics nomogram for discrimination between Brucella spondylitis and Pyogenic spondylitis

Background

Pyogenic spondylitis (PS) and Brucella spondylitis (BS) are commonly seen spinal infectious diseases. Both types can lead to vertebral destruction, kyphosis, and long-term neurological deficits if not promptly diagnosed and treated. Therefore, accurately diagnosis is crucial for personalized therapy. Distinguishing between PS and BS in everyday clinical settings is challenging due to the similarity of their clinical symptoms and imaging features. Hence, this study aims to evaluate the effectiveness of a radiomics nomogram using magnetic resonance imaging (MRI) to accurately differentiate between the two types of spondylitis.

Methods

Clinical and MRI data from 133 patients (2017-2022) with pathologically confirmed PS and BS (68 and 65 patients, respectively) were collected. We have divided patients into training and testing cohorts. In order to develop a clinical diagnostic model, logistic regression was utilized to fit a conventional clinical model (M1). Radiomics features were extracted from sagittal fat-suppressed T2-weighted imaging (FS-T2WI) sequence. The radiomics features were preprocessed, including scaling using Z-score and undergoing univariate analysis to eliminate redundant features. Furthermore, the Least Absolute Shrinkage and Selection Operator (LASSO) was employed to develop a radiomics score (M2). A composite model (M3) was created by combining M1 and M2. Subsequently, calibration and decision curves were generated to evaluate the nomogram's performance in both training and testing groups. The diagnostic performance of each model and the indication was assessed using the receiver operating curve (ROC) with its area under the curve (AUC). Finally, we used the SHapley Additive exPlanations (SHAP) model explanations technique to interpret the model result.

Results

We have finally selected 9 significant features from sagittal FS-T2WI sequences. In the differential diagnosis of PS and BS, the AUC values of M1, M2, and M3 in the testing set were 0.795, 0.859, and 0.868. The composite model exhibited a high degree of concurrence with the ideal outcomes, as evidenced by the calibration curves. The nomogram's possible clinical application values were indicated by the decision curve analysis. By using SHAP values to represent prediction outcomes, our model's prediction results are more understandable.

Conclusions

The implementation of a nomogram that integrates MRI and clinical data has the potential to significantly enhance the accuracy of discriminating between PS and BS within clinical settings.

Radiogenomics: Contemporary Applications in the Management of Rectal Cancer

Radiogenomics, a sub-domain of radiomics, refers to the prediction of underlying tumour biology using non-invasive imaging markers. This novel technology intends to reduce the high costs, workload and invasiveness associated with traditional genetic testing via the development of 'imaging biomarkers' that have the potential to serve as an alternative 'liquid-biopsy' in the determination of tumour biological characteristics. Radiogenomics also harnesses the potential to unlock aspects of tumour biology which are not possible to assess by conventional biopsy-based methods, such as full tumour burden, intra-/inter-lesion heterogeneity and the possibility of providing the information of tumour biology longitudinally. Several studies have shown the feasibility of developing a radiogenomic-based signature to predict treatment outcomes and tumour characteristics; however, many lack prospective, external validation. We performed a systematic review of the current literature surrounding the use of radiogenomics in rectal cancer to predict underlying tumour biology.

Current Status and Future of Artificial Intelligence in MM Imaging: A Systematic Review

Artificial intelligence (AI) has attracted increasing attention as a tool for the detection and management of several medical conditions. Multiple myeloma (MM), a malignancy characterized by uncontrolled proliferation of plasma cells, is one of the most common hematologic malignancies, which relies on imaging for diagnosis and management. We aimed to review the current literature and trends in AI research of MM imaging. This study was performed according to the PRISMA guidelines. Three main concepts were used in the search algorithm, including "artificial intelligence" in "radiologic examinations" of patients with "multiple myeloma". The algorithm was used to search the PubMed, Embase, and Web of Science databases. Articles were screened based on the inclusion and exclusion criteria. In the end, we used the checklist for Artificial Intelligence in Medical Imaging (CLAIM) criteria to evaluate the manuscripts. We provided the percentage of studies that were compliant with each criterion as a measure of the quality of AI research on MM. The initial search yielded 977 results. After reviewing them, 14 final studies were selected. The studies used a wide array of imaging modalities. Radiomics analysis and segmentation tasks were the most popular studies (10/14 studies). The common purposes of radiomics studies included the differentiation of MM bone lesions from other lesions and the prediction of relapse. The goal of the segmentation studies was to develop algorithms for the automatic segmentation of important structures in MM. Dice score was the most common assessment tool in segmentation studies, which ranged from 0.80 to 0.97. These studies show that imaging is a valuable data source for medical AI models and plays an even greater role in the management of MM.

Deep learning-based segmentation of dental implants on cone-beam computed tomography images: A validation study

OBJECTIVES

To train and validate a cloud-based convolutional neural network (CNN) model for automated segmentation (AS) of dental implant and attached prosthetic crown on cone-beam computed tomography (CBCT) images.

METHODS

A total dataset of 280 maxillomandibular jawbone CBCT scans was acquired from patients who underwent implant placement with or without coronal restoration. The dataset was randomly divided into three subsets: training set (n = 225), validation set (n = 25) and testing set (n = 30). A CNN model was developed and trained using expert-based semi-automated segmentation (SS) of the implant and attached prosthetic crown as the ground truth. The performance of AS was assessed by comparing with SS and manually corrected automated segmentation referred to as refined-automated segmentation (R-AS). Evaluation metrics included timing, voxel-wise comparison based on confusion matrix and 3D surface differences.

RESULTS

The average time required for AS was 60 times faster (<30 s) than the SS approach. The CNN model was highly effective in segmenting dental implants both with and without coronal restoration, achieving a high dice similarity coefficient score of 0.92±0.02 and 0.91±0.03, respectively. Moreover, the root mean square deviation values were also found to be low (implant only: 0.08±0.09 mm, implant+restoration: 0.11±0.07 mm) when compared with R-AS, implying high AI segmentation accuracy.

CONCLUSIONS

The proposed cloud-based deep learning tool demonstrated high performance and time-efficient segmentation of implants on CBCT images.

CLINICAL SIGNIFICANCE

AI-based segmentation of implants and prosthetic crowns can minimize the negative impact of artifacts and enhance the generalizability of creating dental virtual models. Furthermore, incorporating the suggested tool into existing CNN models specialized for segmenting anatomical structures can improve pre-surgical planning for implants and post-operative assessment of peri‑implant bone levels.

Fusion-based tensor radiomics using reproducible features: Application to survival prediction in head and neck cancer

BACKGROUND

Numerous features are commonly generated in radiomics applications as applied to medical imaging, and identification of robust radiomics features (RFs) can be an important step to derivation of reliable, reproducible solutions. In this work, we utilize a tensor radiomics (TR) framework, where numerous fusions are explored, to generate different flavours of RFs, and we aimed to identify RFs that are robust to fusion techniques in head and neck cancer. Overall, we aimed to predict progression-free survival (PFS) using Hybrid Machine Learning Systems (HMLS) and reproducible RFs.

METHODS

The study was performed on 408 patients with head and neck cancer from The Cancer Imaging Archive. After image preprocessing, 15 fusion techniques were employed to combine Positron Emission Tomography (PET) and Computed Tomography (CT) images. Subsequently, 215 RFs were extracted through a standardized radiomics software, with 17 'flavours' generated using PET-only, CT-only, and 15 fused PET&CT images. The variability of RFs across flavours was studied using the Intraclass Correlation Coefficient (ICC). Furthermore, the features were categorized into seven reliability groups, 106 reproducible RFs with ICC>0.75 were selected, highly correlated flavours were removed, Principal Component Analysis was used to convert 17 flavours to 1 attribute, the polynomial function was utilized to increase RFs, and Analysis of variance (ANOVA) was used to select the relevant attributes. Finally, 3 classifiers including Random Forest (RFC), Logistic regression (LR), and Multi-layer perceptron were applied to the preselected relevant attributes to predict binary PFS. In 5-fold cross-validation, 80% of 4 divisions were utilized to train the model, and the remaining 20% was utilized to evaluate the model. Further, the remaining fold was used for external nested testing.

RESULTS

Reliability analysis indicated that most morphological features belong to the high-reliability category. By contrast, local intensity and statistical features extracted from images belong to the low-reliability category. In the tensor framework, the highest 5-fold cross-validation accuracy of 76.7%±3.3% with an external nested testing of 70.6%±6.7% resulted from the reproducible TR+polynomial function+ANOVA+LR algorithm while the accuracy of 70.0%±4.2% with the external nested testing of 67.7%±4.9% was achieved through the PCA fusion+RFC (non-tensor paradigm).

CONCLUSIONS

This study demonstrated that using reproducible RFs as utilized within a tensor fusion radiomics framework, linked with ANOVA and LR, added value to prediction of progression-free survival outcome in head and neck cancer patients.

Dynamic contrast-enhanced magnetic resonance imaging radiomics analysis based on intratumoral subregions for predicting luminal and nonluminal breast cancer

Background

Breast cancer is a heterogeneous disease with different morphological and biological characteristics. The molecular subtypes of breast cancer are closely related to the treatment and prognosis of patients. In order to predict the luminal type of breast cancer in a noninvasive manner, our study developed and validated a radiomics nomogram combining clinical factors with a radiomics score based on the features of the intratumoral subregion to distinguish between luminal and nonluminal breast cancer.

Methods

From January 2018 to January 2020, 153 women with clinically and pathologically diagnosed breast cancer with an average age of 50.08 years were retrospectively analyzed. Using a semiautomatic segmentation method, the whole tumor was divided into 3 subregions on the basis of the time required for the contrast agent to reach its peak; 540 features were extracted from 3 subregions and the whole tumor region. Subsequently, 2 machine learning classifiers were developed. The least absolute shrinkage and selection operator method was used for feature selection and radiomics score (Rad-score) construction. Moreover, multivariable logistic regression analysis was applied to select independent factors from the Rad-score and clinical factors to establish a prediction model in the form of a nomogram. The performance of the nomogram was evaluated through calibration, discrimination, and clinical usefulness.

Results

The prediction performance of texture features from the rapid subregion was the best in the 3 intratumoral subregions, and the area under the receiver operating characteristic curve (AUC) values in the training and validation cohort were 0.805 (95% CI: 0.719-0.892) and 0.737 (95% CI: 0.581-0.893), respectively. The Rad-score, consisting of 5 features from the rapid subregion, was associated with the luminal type of breast cancer (P=0.001 and P=0.035 in the training and validation cohorts, respectively). The predictors included in the personalized prediction nomogram included Rad-score, human epidermal growth factor receptor 2 (HER2) status, and tumor histological grade. The nomogram showed good discrimination, with an area under the receiver operating characteristic curve in the training and validation cohorts of 0.830 (95% CI: 0.746-0.896) and 0.879 (95% CI: 0.748-0.957), respectively. The calibration curve of the 2 cohorts and decision curve analysis demonstrated that the nomogram had good calibration and clinical usefulness.

Conclusions

We proposed a nomogram model that combined clinical factors and Rad-score, which showed good performance in predicting the luminal type of breast cancer.

Morphometry-based radiomics for predicting therapeutic response in patients with gliomas following radiotherapy

Introduction

Gliomas are still considered as challenging in oncologic management despite the developments in treatment approaches. The complete elimination of a glioma might not be possible even after a treatment and assessment of therapeutic response is important to determine the future course of actions for patients with such cancers. In the recent years radiomics has emerged as a promising solution with potential applications including prediction of therapeutic response. Hence, this study was focused on investigating whether morphometry-based radiomics signature could be used to predict therapeutic response in patients with gliomas following radiotherapy.

Methods

105 magnetic resonance (MR) images including segmented and non-segmented images were used to extract morphometric features and develop a morphometry-based radiomics signature. After determining the appropriate machine learning algorithm, a prediction model was developed to predict the therapeutic response eliminating the highly correlated features as well as without eliminating the highly correlated features. Then the model performance was evaluated.

Results

Tumor grade had the highest contribution to develop the morphometry-based signature. Random forest provided the highest accuracy to train the prediction model derived from the morphometry-based radiomics signature. An accuracy of 86% and area under the curve (AUC) value of 0.91 were achieved for the prediction model evaluated without eliminating the highly correlated features whereas accuracy and AUC value were 84% and 0.92 respectively for the prediction model evaluated after eliminating the highly correlated features.

Discussion

Nonetheless, the developed morphometry-based radiomics signature could be utilized as a noninvasive biomarker for therapeutic response in patients with gliomas following radiotherapy.

Promoting the application of pediatric radiomics via an integrated medical engineering approach

Radiomics is widely used in adult tumors but has been rarely applied to the field of pediatrics. Promoting the application of radiomics in pediatric diseases, especially in the early diagnosis and stratified treatment of tumors, is of great value to the realization of the WHO 2030 “Global Initiative for Childhood Cancer.” This paper discusses the general characteristics of radiomics, the particularity of its application to pediatric diseases, and the current status and prospects of pediatric radiomics. Radiomics is a data‐driven science, and the combination of medicine and engineering plays a decisive role in improving data quality, data diversity, and sample size. Compared with adult radiomics, pediatric radiomics is significantly different in data type, disease spectrum, disease staging, and progression. Some progress has been made in the identification, classification, stratification, survival prediction, and prognosis of tumor diseases. In the future, big data applications from multiple centers and cross‐talent training should be strengthened to improve the benefits for clinical workers and children.

Radiomics and artificial intelligence for precision medicine in lung cancer treatment

Lung cancer is the leading cause of cancer-related deaths worldwide. It exhibits, at the mesoscopic scale, phenotypic characteristics that are generally indiscernible to the human eye but can be captured non-invasively on medical imaging as radiomic features, which can form a high dimensional data space amenable to machine learning. Radiomic features can be harnessed and used in an artificial intelligence paradigm to risk stratify patients, and predict for histological and molecular findings, and clinical outcome measures, thereby facilitating precision medicine for improving patient care. Compared to tissue sampling-driven approaches, radiomics-based methods are superior for being non-invasive, reproducible, cheaper, and less susceptible to intra-tumoral heterogeneity. This review focuses on the application of radiomics, combined with artificial intelligence, for delivering precision medicine in lung cancer treatment, with discussion centered on pioneering and groundbreaking works, and future research directions in the area.

Radiomics Based on Single-Phase CTA for Distinguishing Left Atrial Appendage Thrombus from Circulatory Stasis in Patients with Atrial Fibrillation before Ablation

Differentiation of left atrial appendage thrombus (LAAT) and left atrial appendage (LAA) circulatory stasis is difficult when based only on single-phase computed tomography angiography (CTA) in routine clinical practice. Radiomics provides a promising tool for their identification. We retrospectively enrolled 204 (training set: 144; test set: 60) atrial fibrillation patients before ablation, including 102 LAAT and 102 circulatory stasis patients. Radiomics software was used to segment whole LAA on single-phase CTA images and extract features. Models were built and compared via a multivariable logistic regression algorithm and area under of the receiver operating characteristic curves (AUCs), respectively. For the radiomics model, radiomics clinical model, radiomics radiological model, and combined model, the AUCs were 0.82, 0.86, 0.90, 0.93 and 0.82, 0.82, 0.84, 0.85 in the training set and the test set, respectively (p < 0.05). One clinical feature (rheumatic heart disease) and four radiological features (transverse diameter of left atrium, volume of left atrium, location of LAA, shape of LAA) were added to the combined model. The combined model exhibited excellent differential diagnostic performances between LAAT and circulatory stasis without increasing extra radiation exposure. The single-phase, CTA-based radiomics analysis shows potential as an effective tool for accurately detecting LAAT in patients with atrial fibrillation before ablation.

Seeing beyond the tumor: computed tomography image-based radiomic analysis helps identify ovarian clear cell carcinoma subtype in epithelial ovarian cancer

OBJECTIVE

To develop and validate a model that can preoperatively identify the ovarian clear cell carcinoma (OCCC) subtype in epithelial ovarian cancer (EOC) using CT imaging radiomics and clinical data.

MATERIAL AND METHODS

We retrospectively analyzed data from 282 patients with EOC (training set = 225, testing set = 57) who underwent pre-surgery CT examinations. Patients were categorized into OCCC or other EOC subtypes based on postoperative pathology. Seven clinical characteristics (age, cancer antigen [CA]-125, CA-199, endometriosis, venous thromboembolism, hypercalcemia, stage) were collected. Primary tumors were manually delineated on portal venous-phase images, and 1218 radiomic features were extracted. The F-test-based feature selection method and logistic regression algorithm were used to build the radiomic signature, clinical model, and integrated model. To explore the effects of integrated model-assisted diagnosis, five radiologists independently interpreted images in the testing set and reevaluated cases two weeks later with knowledge of the integrated model's output. The diagnostic performances of the predictive models, radiologists, and radiologists aided by the integrated model were evaluated.

RESULTS

The integrated model containing the radiomic signature (constructed by four wavelet radiomic features) and three clinical characteristics (CA-125, endometriosis, and hypercalcinemia), showed better diagnostic performance (AUC = 0.863 [0.762-0.964]) than the clinical model (AUC = 0.792 [0.630-0.953], p = 0.295) and the radiomic signature alone (AUC = 0.781 [0.636-0.926], p = 0.185). The diagnostic sensitivities of the radiologists were significantly improved when using the integrated model (p = 0.023-0.041), while the specificities and accuracies were maintained (p = 0.074-1.000).

CONCLUSION

Our integrated model shows great potential to facilitate the early identification of the OCCC subtype in EOC, which may enhance subtype-specific therapy and clinical management.

Radiomics and Radiogenomics in Pelvic Oncology: Current Applications and Future Directions

Radiomics refers to the conversion of medical imaging into high-throughput, quantifiable data in order to analyse disease patterns, guide prognosis and aid decision making. Radiogenomics is an extension of radiomics that combines conventional radiomics techniques with molecular analysis in the form of genomic and transcriptomic data, serving as an alternative to costly, labour-intensive genetic testing. Data on radiomics and radiogenomics in the field of pelvic oncology remain novel concepts in the literature. We aim to perform an up-to-date analysis of current applications of radiomics and radiogenomics in the field of pelvic oncology, particularly focusing on the prediction of survival, recurrence and treatment response. Several studies have applied these concepts to colorectal, urological, gynaecological and sarcomatous diseases, with individual efficacy yet poor reproducibility. This article highlights the current applications of radiomics and radiogenomics in pelvic oncology, as well as the current limitations and future directions. Despite a rapid increase in publications investigating the use of radiomics and radiogenomics in pelvic oncology, the current evidence is limited by poor reproducibility and small datasets. In the era of personalised medicine, this novel field of research has significant potential, particularly for predicting prognosis and guiding therapeutic decisions. Future research may provide fundamental data on how we treat this cohort of patients, with the aim of reducing the exposure of high-risk patients to highly morbid procedures.

Deep versus Handcrafted Tensor Radiomics Features: Prediction of Survival in Head and Neck Cancer Using Machine Learning and Fusion Techniques

BACKGROUND

Although handcrafted radiomics features (RF) are commonly extracted via radiomics software, employing deep features (DF) extracted from deep learning (DL) algorithms merits significant investigation. Moreover, a "tensor'' radiomics paradigm where various flavours of a given feature are generated and explored can provide added value. We aimed to employ conventional and tensor DFs, and compare their outcome prediction performance to conventional and tensor RFs.

METHODS

408 patients with head and neck cancer were selected from TCIA. PET images were first registered to CT, enhanced, normalized, and cropped. We employed 15 image-level fusion techniques (e.g., dual tree complex wavelet transform (DTCWT)) to combine PET and CT images. Subsequently, 215 RFs were extracted from each tumor in 17 images (or flavours) including CT only, PET only, and 15 fused PET-CT images through the standardized-SERA radiomics software. Furthermore, a 3 dimensional autoencoder was used to extract DFs. To predict the binary progression-free-survival-outcome, first, an end-to-end CNN algorithm was employed. Subsequently, we applied conventional and tensor DFs vs. RFs as extracted from each image to three sole classifiers, namely multilayer perceptron (MLP), random-forest, and logistic regression (LR), linked with dimension reduction algorithms.

RESULTS

DTCWT fusion linked with CNN resulted in accuracies of 75.6 ± 7.0% and 63.4 ± 6.7% in five-fold cross-validation and external-nested-testing, respectively. For the tensor RF-framework, polynomial transform algorithms + analysis of variance feature selector (ANOVA) + LR enabled 76.67 ± 3.3% and 70.6 ± 6.7% in the mentioned tests. For the tensor DF framework, PCA + ANOVA + MLP arrived at 87.0 ± 3.5% and 85.3 ± 5.2% in both tests.

CONCLUSIONS

This study showed that tensor DF combined with proper machine learning approaches enhanced survival prediction performance compared to conventional DF, tensor and conventional RF, and end-to-end CNN frameworks.

Assessment of Variabilities in Lung-Contouring Methods on CBCT Preclinical Radiomics Outputs

Simple Summary

This study is the first to evaluate the impact of contouring differences on radiomics analysis in preclinical CBCT scans. We found that the variation in quantitative image readouts was greater between segmentation tools than between observers.

Abstract

Radiomics image analysis has the potential to uncover disease characteristics for the development of predictive signatures and personalised radiotherapy treatment. Inter-observer and inter-software delineation variabilities are known to have downstream effects on radiomics features, reducing the reliability of the analysis. The purpose of this study was to investigate the impact of these variabilities on radiomics outputs from preclinical cone-beam computed tomography (CBCT) scans. Inter-observer variabilities were assessed using manual and semi-automated contours of mouse lungs (n = 16). Inter-software variabilities were determined between two tools (3D Slicer and ITK-SNAP). The contours were compared using Dice similarity coefficient (DSC) scores and the 95th percentile of the Hausdorff distance (HD_95p) metrics. The good reliability of the radiomics outputs was defined using intraclass correlation coefficients (ICC) and their 95% confidence intervals. The median DSC scores were high (0.82–0.94), and the HD_95p metrics were within the submillimetre range for all comparisons. the shape and NGTDM features were impacted the most. Manual contours had the most reliable features (73%), followed by semi-automated (66%) and inter-software (51%) variabilities. From a total of 842 features, 314 robust features overlapped across all contouring methodologies. In addition, our results have a 70% overlap with features identified from clinical inter-observer studies.

Prediction of the mitotic index and preoperative risk stratification of gastrointestinal stromal tumors with CT radiomic features

OBJECTIVE

The objective is to develop a mitotic prediction model and preoperative risk stratification nomogram for gastrointestinal stromal tumor (GIST) based on computed tomography (CT) radiomic features.

METHODS

A total of 267 GIST patients from 2009.07 to 2015.09 were retrospectively collected and randomly divided into (6:4) training cohort and validation cohort. The 2D-tumor region of interest was delineated from the portal-phase images on contrast-enhanced (CE)-CT, and radiomic features were extracted. Lasso regression method was used to select valuable features to establish a radiomic model for predicting mitotic index in GIST. Finally, the nomogram of preoperative risk stratification was constructed by combining the radiomic features and clinical risk factors.

RESULTS

Four radiomic features closely related to the level of mitosis were obtained, and a mitotic radiomic model was constructed. The area under the curve (AUC) of the radiomics signature model used to predict mitotic levels in training and validation cohorts (training cohort AUC = 0.752; 95% confidence interval [95%CI] 0.674-0.829; validation cohort AUC = 0.764; 95% CI 0.667-0.862). Finally, the preoperative risk stratification nomogram combining radiomic features was equivalent to the clinically recognized gold standard AUC (0.965 vs. 0.983) (p = 0.117). The Cox regression analysis found that the nomogram score was one of the independent risk factors for the long-term prognosis of the patients.

CONCLUSION

Preoperative CT radiomic features can effectively predict the level of mitosis in GIST, and combined with preoperative tumor size, accurate preoperative risk stratification can be performed to guide clinical decision-making and individualized treatment.

A pilot study of MRI radiomics for high-risk prostate cancer stratification in 1.5 T MR-guided radiotherapy

PURPOSE

To investigate the potential value of MRI radiomics obtained from a 1.5 T MRI-guided linear accelerator (MR-LINAC) for D'Amico high-risk prostate cancer (PC) classification in MR-guided radiotherapy (MRgRT).

METHODS

One hundred seventy-six consecutive PC patients underwent 1.5 T MRgRT treatment were retrospectively enrolled. Each patient received one or two pretreatment T₂ -weighted MRI scans on a 1.5 T MR-LINAC. The endpoint was to differentiate high-risk from low/intermediate-risk PC based on D'Amico criteria using MRI-radiomics. Totally 1023 features were extracted from clinical target volume (CTV) and planning target volume (PTV). Intraclass correlation coefficient of scan-rescan repeatability, feature correlation, and recursive feature elimination were used for feature dimension reduction. Least absolute shrinkage and selection operator regression was employed for model construction. Receiver operating characteristic area under the curve (AUC) analysis was used for model performance assessment in both training and testing data.

RESULTS

One hundred and eleven patients fulfilled all criteria were finally included: 76 for training and 35 for testing. The constructed MRI-radiomics models extracted from CTV and PTV achieved the AUC of 0.812 and 0.867 in the training data, without significant difference (P = 0.083). The model performances remained in the testing. The sensitivity, specificity, and accuracy were 85.71%, 64.29%, and 77.14% for the PTV-based model; and 71.43%, 71.43%, and 71.43% for the CTV-based model. The corresponding AUCs were 0.718 and 0.750 (P = 0.091) for CTV- and PTV-based models.

CONCLUSION

MRI-radiomics obtained from a 1.5 T MR-LINAC showed promising results in D'Amico high-risk PC stratification, potentially helpful for the future PC MRgRT. Prospective studies with larger sample sizes and external validation are warranted for further verification.

Repeatability of radiomics studies in colorectal cancer: a systematic review

BACKGROUND

Recently, radiomics has been widely used in colorectal cancer, but many variable factors affect the repeatability of radiomics research. This review aims to analyze the repeatability of radiomics studies in colorectal cancer and to evaluate the current status of radiomics in the field of colorectal cancer.

METHODS

The included studies in this review by searching from the PubMed and Embase databases. Then each study in our review was evaluated using the Radiomics Quality Score (RQS). We analyzed the factors that may affect the repeatability in the radiomics workflow and discussed the repeatability of the included studies.

RESULTS

A total of 188 studies was included in this review, of which only two (2/188, 1.06%) studies controlled the influence of individual factors. In addition, the median score of RQS was 11 (out of 36), range-1 to 27.

CONCLUSIONS

The RQS score was moderately low, and most studies did not consider the repeatability of radiomics features, especially in terms of Intra-individual, scanners, and scanning parameters. To improve the generalization of the radiomics model, it is necessary to further control the variable factors of repeatability.

Contrast-enhanced CT radiomics improves the prediction of abdominal aortic aneurysm progression

OBJECTIVES

To determine if three-dimensional (3D) radiomic features of contrast-enhanced CT (CECT) images improve prediction of rapid abdominal aortic aneurysm (AAA) growth.

METHODS

This longitudinal cohort study retrospectively analyzed 195 consecutive patients (mean age, 72.4 years ± 9.1) with a baseline CECT and a subsequent CT or MR at least 6 months later. 3D radiomic features were measured for 3 regions of the AAA, viz. the vessel lumen only; the intraluminal thrombus (ILT) and aortic wall only; and the entire AAA sac (lumen, ILT, and wall). Multiple machine learning (ML) models to predict rapid growth, defined as the upper tercile of observed growth (> 0.25 cm/year), were developed using data from 60% of the patients. Diagnostic accuracy was evaluated using the area under the receiver operating characteristic curve (AUC) in the remaining 40% of patients.

RESULTS

The median AAA maximum diameter was 3.9 cm (interquartile range [IQR], 3.3-4.4 cm) at baseline and 4.4 cm (IQR, 3.7-5.4 cm) at the mean follow-up time of 3.2 ± 2.4 years (range, 0.5-9 years). A logistic regression model using 7 radiomic features of the ILT and wall had the highest AUC (0.83; 95% confidence interval [CI], 0.73-0.88) in the development cohort. In the independent test cohort, this model had a statistically significantly higher AUC than a model including maximum diameter, AAA volume, and relevant clinical factors (AUC = 0.78, 95% CI, 0.67-0.87 vs AUC = 0.69, 95% CI, 0.57-0.79; p = 0.04).

CONCLUSION

A radiomics-based method focused on the ILT and wall improved prediction of rapid AAA growth from CECT imaging.

KEY POINTS

• Radiomic analysis of 195 abdominal CECT revealed that an ML-based model that included textural features of intraluminal thrombus (if present) and aortic wall improved prediction of rapid AAA progression compared to maximum diameter. • Predictive accuracy was higher when radiomic features were obtained from the thrombus and wall as opposed to the entire AAA sac (including lumen), or the lumen alone. • Logistic regression of selected radiomic features yielded similar accuracy to predict rapid AAA progression as random forests or support vector machines.

Whole-Lesion Apparent Diffusion Coefficient Histogram Analysis for Assessing Normal-Sized Lymph Node Metastasis in Cervical Cancer: Comparison Between Readout-Segmented and Single-Shot Echo-Planar Diffusion-Weighted Imaging

OBJECTIVE

To compare the value of whole-lesion apparent diffusion coefficient (ADC) histogram analysis derived from readout-segmented echo-planar imaging (RS-EPI) and single-shot echo-planar imaging (SS-EPI) diffusion-weighted imaging (DWI) in evaluating normal-sized lymph node metastasis (LNM) in cervical cancer.

METHODS

Seventy-six pathologically confirmed cervical cancer patients (stages IB and IIA) were enrolled, including 61 patients with non-LNM (group A) and 15 patients with normal-sized LNM (group B). The recorded tumor volume on T2-weighted imaging was the reference against which both DWIs were evaluated. Each ADC histogram parameter (including ADCmax, ADC90, ADCmedian, ADCmean, ADC10, ADCmin, ADCskewness, ADCkurtosis, and ADCentropy) was compared between SS-EPI and RS-EPI and between the 2 groups.

RESULTS

There was no significant difference in tumor volume between the 2 DWIs and T2-weighted imaging (both P > 0.05). Higher ADCmax and ADCentropy but lower ADC10, ADCmin and ADCskewness were found in SS-EPI than those in RS-EPI (all P < 0.05). For SS-EPI, lower ADC90 and higher ADCkurtosis were found in group B than those in group A (both P < 0.05). For RS-EPI, lower ADC90 and higher ADCkurtosis and ADCentropy were found in group B than those in group A (all P < 0.05). Readout-segmented echo-planar imaging ADCkurtosis showed the highest area under the curve of 0.792 in the differentiation of the 2 groups (sensitivity, 80%; specificity, 73.77%).

CONCLUSIONS

Compared with SS-EPI, the ADC histogram parameters derived from RS-EPI were more accurate, and ADCkurtosis held great potential in differentiating normal-sized LNM in cervical cancer.

Machine learning for predicting accuracy of lung and liver tumor motion tracking using radiomic features

Background

Internal tumor motion is commonly predicted using external respiratory signals. However, the internal/external correlation is complex and patient-specific. The purpose of this study was to develop various models based on the radiomic features of computed tomography (CT) images to predict the accuracy of tumor motion tracking using external surrogates and to find accurate and reliable tracking algorithms.

Methods

Images obtained from a total of 108 and 71 patients pathologically diagnosed with lung and liver cancers, respectively, were examined. Real-time position monitoring motion was fitted to tumor motion, and samples with fitting errors greater than 2 mm were considered positive. Radiomic features were extracted from internal target volumes of average intensity projections, and cross-validation least absolute shrinkage and selection operator (LassoCV) was used to conduct feature selection. Based on the radiomic features, a total of 26 separate models (13 for the lung and 13 for the liver) were trained and tested. Area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were used to assess performance. Relative standard deviation was used to assess stability.

Results

Thirty-three and 22 radiomic features were selected for the lung and liver, respectively. For the lung, the AUC varied from 0.848 (decision tree) to 0.941 [support vector classifier (SVC), logistic regression]; sensitivity varied from 0.723 (extreme gradient boosting) to 0.848 [linear support vector classifier (linearSVC)]; specificity varied from 0.834 (gaussian naive bayes) to 0.936 [multilayer perceptron (MLP), wide and deep (W&D)]; and MLP and W&D had better performance and stability than the median. For the liver, the AUC varied from 0.677 [light gradient boosting machine (Light)] to 0.892 (logistic regression); sensitivity varied from 0.717 (W&D) to 0.862 (MLP); specificity varied from 0.566 (Light) to 0.829 (linearSVC); and logistic regression, MLP, and SVC had better performance and stability than the median.

Conclusions

Respiratory-sensitive radiomic features extracted from CT images of lung and liver tumors were proved to contain sufficient information to establish an external/internal motion relationship. We developed a rapid and accurate method based on radiomics to classify the accuracy of monitoring a patient's external surface for lung and liver tumor tracking. Several machine learning algorithms-in particular, MLP-demonstrated excellent classification performance and stability.

Diagnosis of Acute Aortic Syndromes on Non-Contrast CT Images with Radiomics-Based Machine Learning

We aimed to detect acute aortic syndromes (AAS) on non-contrast computed tomography (NCCT) images using a radiomics-based machine learning model. A total of 325 patients who underwent aortic CT angiography (CTA) were enrolled retrospectively from 2 medical centers in China to form the internal cohort (230 patients, 60 patients with AAS) and the external testing cohort (95 patients with AAS). The internal cohort was divided into the training cohort (n = 135), validation cohort (n = 49), and internal testing cohort (n = 46). The aortic mask was manually delineated on NCCT by a radiologist. Least Absolute Shrinkage and Selection Operator regression (LASSO) was used to filter out nine feature parameters; the Support Vector Machine (SVM) model showed the best performance. In the training and validation cohorts, the SVM model had an area under the curve (AUC) of 0.993 (95% CI, 0.965–1); accuracy (ACC), 0.946 (95% CI, 0.877–1); sensitivity, 0.9 (95% CI, 0.696–1); and specificity, 0.964 (95% CI, 0.903–1). In the internal testing cohort, the SVM model had an AUC of 0.997 (95% CI, 0.992–1); ACC, 0.957 (95% CI, 0.945–0.988); sensitivity, 0.889 (95% CI, 0.888–0.889); and specificity, 0.973 (95% CI, 0.959–1). In the external testing cohort, the ACC was 0.991 (95% CI, 0.937–1). This model can detect AAS on NCCT, reducing misdiagnosis and improving examinations and prognosis.

Gadoxetic Acid-Enhanced MRI-Based Radiomics Signature: A Potential Imaging Biomarker for Identifying Cytokeratin 19-Positive Hepatocellular Carcinoma

Purpose

One subtype of hepatocellular carcinoma (HCC), with cytokeratin 19 expression (CK19+), has shown to be more aggressive and has a poor prognosis. However, CK19+ is determined by immunohistochemical examination using a surgically resected specimen. This study is aimed at establishing a radiomics signature based on preoperative gadoxetic acid-enhanced MRI for predicting CK19 status in HCC. Patients and Methods. Clinicopathological and imaging data were retrospectively collected from patients who underwent hepatectomy between February 2015 and December 2020. Patients who underwent gadoxetic acid-enhanced MRI and had CK19 results of histopathological examination were included. Radiomics features of the manually segmented lesion during the arterial, portal venous, and hepatobiliary phases were extracted. The 10 most reproducible and robust features at each phase were selected for construction of radiomics signatures, and their performance was evaluated by analyzing the area under the curve (AUC). The goodness of fit of the model was assessed by the Hosmer-Lemeshow test.

Results

A total of 110 patients were included. The incidence of CK19(+) HCC was 17% (19/110). Alpha fetoprotein was the only significant clinicopathological variable different between CK19(-) and CK19(+) groups. A majority of the selected radiomics features were wavelet filter-derived features. The AUCs of the three radiomics signatures based on arterial, portal venous, and hepatobiliary phases were 0.70 (95% CI: 0.56-0.83), 0.83 (95% CI: 0.73-0.92), and 0.89 (95% CI: 0.82-0.96), respectively. The three radiomics signatures were integrated, and the fusion signature yielded an AUC of 0.92 (95% CI: 0.86-0.98) and was used as the final model for CK19(+) prediction. The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of the fusion signature was 0.84, 0.89, 0.88, 0.62, and 0.96, respectively. The Hosmer-Lemeshow test showed a good fit of the fusion signature (p > 0.05).

Conclusion

The established radiomics signature based on preoperative gadoxetic acid-enhanced MRI could be an accurate and potential imaging biomarker for HCC CK19(+) prediction.

Contrast enhanced magnetic resonance imaging-based radiomics nomogram for preoperatively predicting expression status of Ki-67 in meningioma: a two-center study

Background

The aim of this study was to develop and validate a radiomics nomogram for preoperative prediction of Ki-67 proliferative index (Ki-67 PI) expression in patients with meningioma.

Methods

A total of 280 patients from 2 independent hospital centers were enrolled. Patients from center I were randomly divided into a training cohort of 168 patients and a test cohort of 72 patients, and 40 patients from center II served as an external validation cohort. Interoperator reproducibility test, Z-score standardization, analysis of variance (ANOVA), and least absolute shrinkage and selection operator (LASSO) binary logistic regression were used to select radiomics features, which were extracted from contrast-enhanced T1-weighted imaging (CE-T1WI) imaging. The radiomics signature for predicting Ki-67 PI expression was developed and validated using 4 classifiers including logistic regression (LR), decision tree (DT), support vector machine (SVM), and adaptive boost (AdaBoost). Finally, combined radiological characteristics with radiomics signature were used to establish the nomogram to predict the risk of high Ki-67 PI expression in patients with meningioma.

Results

Fourteen radiomics features were used to construct the radiomics signature. The radiomics nomogram that incorporated the radiomics signature and radiological characteristics showed excellent discrimination in the training, test, and validation cohorts with areas under the curve of 0.817 (95% CI: 0.753-0.881), 0.822 (95% CI: 0.727-0.916), and 0.845 (95% CI: 0.708-0.982), respectively. In addition, the calibration curve for the nomogram demonstrated good agreement between prediction and actual observation.

Conclusions

The proposed contrast enhanced magnetic resonance imaging (MRI)-based radiomics nomogram could be an effective tool to predict the risk of Ki-67 high expression in patients with meningioma.

Determining the reliable feature change in longitudinal radiomics studies: A methodological approach using the reliable change index

PURPOSE

Determination of reliable change of radiomics feature over time is essential and vital in delta-radiomics, but has not yet been rigorously examined. This study attempts to propose a methodological approach using reliable change index (RCI), a statistical metric to determine the reliability of quantitative biomarker changes by accounting for the baseline measurement standard error, in delta-radiomics. The use of RCI was demonstrated with the MRI data acquired from a group of prostate cancer (PCa) patients treated by 1.5 T MRI-guided radiotherapy (MRgRT).

METHODS

Fifty consecutive PCa patients who underwent five-fractionated MRgRT were retrospectively included, and 1023 radiomics features were extracted from the clinical target volume (CTV) and planning target volume (PTV). The two MRI datasets acquired at the first fraction (MRI11 and MRI21) were used to calculate the baseline feature reliability against image acquisition using intraclass correlation coefficient (ICC). The RCI was constructed based on the baseline feature measurement standard deviation, ICC, and feature value differences at two time points between the fifth (MRI51) and the first fraction MRI (MRI11). The reliable change of features was determined in each patient only if the calculated RCI was over 1.96 or smaller than -1.96. The feature changes between MRI51 and MRI11 were correlated to two patient-reported quality-of-life clinical endpoints of urinary domain summary score (UDSS) and bowel domain summary score (BDSS) in 35 patients using the Spearman correlation test. Only the significant correlations between a feature that was reliably changed in ≥7 patients (20%) by RCI and an endpoint were considered as true significant correlations.

RESULTS

The 352 (34.4%) and 386 (37.7%) features among all 1023 features were determined by RCI to be reliably changed in more than five (10%) patients in the CTV and PTV, respectively. Nineteen features were found reliably changed in the CTV and 31 features in the PTV, respectively, in 10 (20%) or more patients. These features were not necessarily associated with significantly different longitudinal feature values (group p-value < 0.05). Most reliably changed features in more than 10 patients had excellent or good baseline test-retest reliability ICC, while none showed poor reliability. The RCI method ruled out the features to be reliably changed when substantial feature measurement bias was presented. After applying the RCI criterion, only four and five true significant correlations were confirmed with UDSS and BDSS in the CTV, respectively, with low true significance correlation rates of 10.8% (4/37) and 17.9% (5/28). No true significant correlations were found in the PTV.

CONCLUSIONS

The RCI method was proposed for delta-radiomics and demonstrated using PCa MRgRT data. The RCI has advantages over some other statistical metrics commonly used in the previous delta-radiomics studies, and is useful to reliably identify the longitudinal radiomics feature change on an individual basis. This proposed RCI method should be helpful for the development of essential feature selection methodology in delta-radiomics.

Changes in CT-Based Morphological Features of the Kidney with Declining Glomerular Filtration Rate in Chronic Kidney Disease

Chronic kidney disease (CKD) progression involves morphological changes in the kidney, such as decreased length and thickness, with associated histopathological alterations. However, the relationship between morphological changes in the kidneys and glomerular filtration rate (GFR) has not been quantitatively and comprehensively evaluated. We evaluated the three-dimensional size and shape of the kidney using computed tomography (CT)-derived features in relation to kidney function. We included 257 patients aged ≥18 years who underwent non-contrast abdominal CT at the Inha University Hospital. The features were quantified using predefined algorithms in the pyRadiomics package after kidney segmentation. All features, except for flatness, significantly correlated with estimated GFR (eGFR). The surface-area-to-volume ratio (SVR) showed the strongest negative correlation (r = -0.75, p < 0.0001). Kidney size features, such as volume and diameter, showed moderate to high positive correlations; other morphological features showed low to moderate correlations. The calculated area under the receiver operating characteristic (ROC) curve (AUC) for different features ranged from 0.51 (for elongation) to 0.86 (for SVR) for different eGFR thresholds. Diabetes patients had weaker correlations between the studied features and eGFR and showed less bumpy surfaces in three-dimensional visualization. We identified alterations in the CKD kidney based on various three-dimensional shape and size features, with their potential diagnostic value.

Precision (repeatability and reproducibility) of papillary and peripapillary vascular density measurements using optical coherence tomography angiography in children

Importance

Optical coherence tomography angiography (OCTA) has been widely applied into children, however, few studies have assessed the repeatability and reproducibility of papillary and peripapillary VD in healthy children.

Objective

To assess the precision of papillary and peripapillary vascular density (VD) measurements using optical coherence tomography angiography (OCTA) and analyze the effects of the signal strength index (SSI) and axial length (AL) on precision estimates.

Design setting and participants

This was a prospective observational study. Seventy-eight children aged 6-16 years underwent 4.5 × 4.5 mm OCTA (RTVue XR Avanti) disc scans: two scans by one examiner (repeatability) and two additional scans by another examiner (reproducibility). Within-subject standard deviation (Sw), test-retest reproducibility (TRT), within-subject coefficient of variation (CoV), intraclass correlation coefficient (ICC), and Bland-Altman analysis were performed.

Main outcomes and measures

In repeatability measurement, the fluctuation ranges (minimum to maximum) of VD between intraexaminer A/B in Sw, TRT, CoV, and ICC were (1.05-2.17)% / (1.16-2.32)%, (2.9-6)% / (3.21-6.44)%, (1.9-4.47)% / (2.08-5)%, and (0.588-0.783)% / (0.633-0.803)%, respectively. In reproducibility measurement, the fluctuation ranges of VD in Sw, TRT, CoV, and ICC were 1.11-2.13%, 3.07-5.91%, 1.99-4.41%, and 0.644-0.777%, respectively. VD was negatively correlated with SSI in most sectors of the peripapillary (e.g., inferior nasal, temporal inferior, temporal superior, superior temporal, and superior nasal). AL was positively correlated with inferior temporal VD and negatively correlated with superior nasal VD.

Conclusion and relevance

Optical coherence tomography angiography showed moderate-to-good repeatability and reproducibility for papillary and peripapillary perfusion measurements in healthy children. The SSI value affects most of the peripapillary VD, while AL affects only the temporal inferior and nasal superior peripapillary VD.

Joint EANM/SNMMI guideline on radiomics in nuclear medicine : Jointly supported by the EANM Physics Committee and the SNMMI Physics, Instrumentation and Data Sciences Council

PURPOSE

The purpose of this guideline is to provide comprehensive information on best practices for robust radiomics analyses for both hand-crafted and deep learning-based approaches.

METHODS

In a cooperative effort between the EANM and SNMMI, we agreed upon current best practices and recommendations for relevant aspects of radiomics analyses, including study design, quality assurance, data collection, impact of acquisition and reconstruction, detection and segmentation, feature standardization and implementation, as well as appropriate modelling schemes, model evaluation, and interpretation. We also offer an outlook for future perspectives.

CONCLUSION

Radiomics is a very quickly evolving field of research. The present guideline focused on established findings as well as recommendations based on the state of the art. Though this guideline recognizes both hand-crafted and deep learning-based radiomics approaches, it primarily focuses on the former as this field is more mature. This guideline will be updated once more studies and results have contributed to improved consensus regarding the application of deep learning methods for radiomics. Although methodological recommendations in the present document are valid for most medical image modalities, we focus here on nuclear medicine, and specific recommendations when necessary are made for PET/CT, PET/MR, and quantitative SPECT.

Magnetic resonance-based radiomics nomogram for predicting microsatellite instability status in endometrial cancer

Background

Microsatellite instability (MSI) status is an important indicator for screening patients with endometrial cancer (EC) who have potential Lynch syndrome (LS) and may benefit from immunotherapy. This study aimed to develop a magnetic resonance imaging (MRI)-based radiomics nomogram for the prediction of MSI status in EC.

Methods

A total of 296 patients with histopathologically diagnosed EC were enrolled, and their MSI status was determined using immunohistochemical (IHC) analysis. Patients were randomly divided into the training cohort (n=236) and the validation cohort (n=60) at a ratio of 8:2. To predict the MSI status in EC, the tumor radiomics features were extracted from T2-weighted images and contrast-enhanced T1-weighted images, which in turn were selected using one-way analysis of variance (ANOVA) and the least absolute shrinkage and selection operator (LASSO) algorithm to build the radiomics signature (radiomics score; radscore) model. Five clinicopathologic characteristics were used to construct a clinicopathologic model. Finally, the nomogram model combining radscore and clinicopathologic characteristics was constructed. The performance of the three models was evaluated using receiver operating characteristic (ROC), calibration, and decision curve analyses (DCA).

Results

Totals of 21 radiomics features and five clinicopathologic characteristics were selected to develop the radscore and clinicopathological models. The radscore and clinicopathologic models achieved an area under the curve (AUC) of 0.752 and 0.600, respectively, in the training cohort; and of 0.723 and 0.615, respectively, in the validation cohort. The radiomics nomogram model showed improved discrimination efficiency compared with the radscore and clinicopathologic models, with an AUC of 0.773 and 0.740 in the training and validation cohorts, respectively. The calibration curve analysis and DCA showed favorable calibration and clinical utility of the nomogram model.

Conclusions

The nomogram incorporating MRI-based radiomics features and clinicopathologic characteristics could be a potential tool for the prediction of MSI status in EC.

Development of a novel nomogram-based model incorporating 3D radiomic signatures and lung CT radiological features for differentiating invasive adenocarcinoma from adenocarcinoma in situ and minimally invasive adenocarcinoma

Background

Lung cancer is one of the most serious cancers in the world. Subtypes of lung adenocarcinoma can be quickly distinguished by analyzing 3D radiomic signatures and radiological features.

Methods

This study included 493 patients from 3 hospitals with a total of 506 lesions confirmed as minimally invasive adenocarcinoma (MIA), adenocarcinoma in situ (AIS), or invasive adenocarcinoma (IAC). After segmenting the lesion area, 3D radiomic signatures were extracted using the PyRadiomics package v. 3.0.1 implemented in Python (https://pyradiomics.readthedocs.io/en/latest/index.html), and the corresponding radiological features were collected. Subsequently, the top 100 features were identified by feature screening methods, including the Spearman rank correlation and minimum redundancy maximum relevance (mRMR) feature selection, and the top 10 features were determined by the least absolute shrinkage and selection operator (LASSO) classifier. Multivariable logistic regression analysis was used to develop a nomogram incorporating 3D radiomic signatures and radiological features in the prediction system. The nomogram was evaluated from multiple perspectives and tested on the validation cohort.

Results

The model combined 3 radiological features and seven 3D radiomic signatures. The area under the curve (AUC) of the model was 0.877 (95% CI: 0.829-0.925) in the training cohort, 0.864 (95% CI: 0.789-0.940) in the testing cohort, and 0.836 (95% CI: 0.749-0.924) in the validation cohort. The nomogram applied in all 3 cohorts showed reliable accuracy and calibration. The decision curve also demonstrated the clinical effectiveness of the nomogram.

Conclusions

In this study, a nomogram-based model combining 3D radiomic signatures and radiological features was developed. Its performance in identifying IAC and MIA/AIS was satisfactory and had clinical value.

Current Applications of Deep Learning and Radiomics on CT and CBCT for Maxillofacial Diseases

The increasing use of computed tomography (CT) and cone beam computed tomography (CBCT) in oral and maxillofacial imaging has driven the development of deep learning and radiomics applications to assist clinicians in early diagnosis, accurate prognosis prediction, and efficient treatment planning of maxillofacial diseases. This narrative review aimed to provide an up-to-date overview of the current applications of deep learning and radiomics on CT and CBCT for the diagnosis and management of maxillofacial diseases. Based on current evidence, a wide range of deep learning models on CT/CBCT images have been developed for automatic diagnosis, segmentation, and classification of jaw cysts and tumors, cervical lymph node metastasis, salivary gland diseases, temporomandibular (TMJ) disorders, maxillary sinus pathologies, mandibular fractures, and dentomaxillofacial deformities, while CT-/CBCT-derived radiomics applications mainly focused on occult lymph node metastasis in patients with oral cancer, malignant salivary gland tumors, and TMJ osteoarthritis. Most of these models showed high performance, and some of them even outperformed human experts. The models with performance on par with human experts have the potential to serve as clinically practicable tools to achieve the earliest possible diagnosis and treatment, leading to a more precise and personalized approach for the management of maxillofacial diseases. Challenges and issues, including the lack of the generalizability and explainability of deep learning models and the uncertainty in the reproducibility and stability of radiomic features, should be overcome to gain the trust of patients, providers, and healthcare organizers for daily clinical use of these models.

Integration of Clinical and CT-Based Radiomic Features for Pretreatment Prediction of Pathologic Complete Response to Neoadjuvant Systemic Therapy in Breast Cancer

The purpose of the present study was to examine the potential of a machine learning model with integrated clinical and CT-based radiomics features in predicting pathologic complete response (pCR) to neoadjuvant systemic therapy (NST) in breast cancer. Contrast-enhanced CT was performed in 329 patients with breast tumors (n = 331) before NST. Pyradiomics was used for feature extraction, and 107 features of seven classes were extracted. Feature selection was performed on the basis of the intraclass correlation coefficient (ICC), and six ICC thresholds (0.7−0.95) were examined to identify the feature set resulting in optimal model performance. Clinical factors, such as age, clinical stage, cancer cell type, and cell surface receptors, were used for prediction. We tried six machine learning algorithms, and clinical, radiomics, and clinical−radiomics models were trained for each algorithm. Radiomics and clinical−radiomics models with gray level co-occurrence matrix (GLCM) features only were also built for comparison. The linear support vector machine (SVM) regression model trained with radiomics features of ICC ≥0.85 in combination with clinical factors performed the best (AUC = 0.87). The performance of the clinical and radiomics linear SVM models showed statistically significant difference after correction for multiple comparisons (AUC = 0.69 vs. 0.78; p < 0.001). The AUC of the radiomics model trained with GLCM features was significantly lower than that of the radiomics model trained with all seven classes of radiomics features (AUC = 0.85 vs. 0.87; p = 0.011). Integration of clinical and CT-based radiomics features was helpful in the pretreatment prediction of pCR to NST in breast cancer.

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.

Predictive performance of radiomic models based on features extracted from pretrained deep networks

OBJECTIVES

In radiomics, generic texture and morphological features are often used for modeling. Recently, features extracted from pretrained deep networks have been used as an alternative. However, extracting deep features involves several decisions, and it is unclear how these affect the resulting models. Therefore, in this study, we considered the influence of such choices on the predictive performance.

METHODS

On ten publicly available radiomic datasets, models were trained using feature sets that differed in terms of the utilized network architecture, the layer of feature extraction, the used set of slices, the use of segmentation, and the aggregation method. The influence of these choices on the predictive performance was measured using a linear mixed model. In addition, models with generic features were trained and compared in terms of predictive performance and correlation.

RESULTS

No single choice consistently led to the best-performing models. In the mixed model, the choice of architecture (AUC + 0.016; p < 0.001), the level of feature extraction (AUC + 0.016; p < 0.001), and using all slices (AUC + 0.023; p < 0.001) were highly significant; using the segmentation had a lower influence (AUC + 0.011; p = 0.023), while the aggregation method was insignificant (p = 0.774). Models based on deep features were not significantly better than those based on generic features (p > 0.05 on all datasets). Deep feature sets correlated moderately with each other (r = 0.4), in contrast to generic feature sets (r = 0.89).

CONCLUSIONS

Different choices have a significant effect on the predictive performance of the resulting models; however, for the highest performance, these choices should be optimized during cross-validation.

A methodological framework for AI-assisted diagnosis of active aortitis using radiomic analysis of FDG PET-CT images: Initial analysis

BACKGROUND

The aim of this study was to explore the feasibility of assisted diagnosis of active (peri-)aortitis using radiomic imaging biomarkers derived from [18F]-Fluorodeoxyglucose Positron Emission Tomography-Computed Tomography (FDG PET-CT) images.

METHODS

The aorta was manually segmented on FDG PET-CT in 50 patients with aortitis and 25 controls. Radiomic features (RF) (n = 107), including SUV (Standardized Uptake Value) metrics, were extracted from the segmented data and harmonized using the ComBat technique. Individual RFs and groups of RFs (i.e., signatures) were used as input in Machine Learning classifiers. The diagnostic utility of these classifiers was evaluated with area under the receiver operating characteristic curve (AUC) and accuracy using the clinical diagnosis as the ground truth.

RESULTS

Several RFs had high accuracy, 84% to 86%, and AUC scores 0.83 to 0.97 when used individually. Radiomic signatures performed similarly, AUC 0.80 to 1.00.

CONCLUSION

A methodological framework for a radiomic-based approach to support diagnosis of aortitis was outlined. Selected RFs, individually or in combination, showed similar performance to the current standard of qualitative assessment in terms of AUC for identifying active aortitis. This framework could support development of a clinical decision-making tool for a more objective and standardized assessment of aortitis.

Harmonization Strategies in Multicenter MRI-Based Radiomics

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

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.