Performance analysis of 2D and 3D image features for computer-assisted speech diagnosis of dental sibilants in Polish children

BACKGROUND AND OBJECTIVE

Sigmatism is a speech disorder concerning sibilants, and its diagnosis affects many Polish children of preschool age. The success of therapy often depends on early and accurate diagnosis. This paper presents research findings on using 2D and 3D (time-related) visual features to analyze the place of articulation, sibilance (the character of a gap between teeth that allows the articulation of sibilant sounds), and tongue positioning in four of twelve Polish sibilants:/s/,/z/,/ʦ/, and/dz/.

METHODS

A dedicated data acquisition system captured the stereovision stream during the speech therapy examination (201 speakers aged 4-8). The material contains 23 words and four logatomes. This study introduces 3D texture and shape features extracted for the mouth, lips, and tongue. The third dimension is the time of articulation, and the volumes reflect the movements of speech organs. The research compares the usability of 3D mode to a 2D approach (mouth texture features; mouth, lips, and tongue shape parameters) described in previous works. The statistical analysis includes Mann-Whitney U test to indicate the significant differences between selected articulation patterns for each sibilant and pronunciation aspect (considering p<0.05).

RESULTS

Overall outcomes suggest the dominance of 3D time-related statistically significant features, especially describing the shape of a tongue. Analysis considering features with at least medium effect size showed that 3D features differentiate dental and interdental articulation in case of/s/,/z/, and/ʦ/, while in case of/dz/ significant parameters were 2D. The 3D mode prevails also in terms of sibilance: analysis of sounds/z/ and/ʦ/ results in 3D features only, but for/s/ and/dz/ outcomes include both 3D and 2D parameters. Analysis of the tongue positioning during articulation in terms of at least moderate effect size suggests a presence of features only in the case of affricates:/ʦ/ (3D features) and/dz/ (2D features). All parameters with at least medium effect size describe the shape of the tongue.

CONCLUSIONS

This research proves the potential of visual data in building computer-aided speech diagnosis systems using non-contact recording tools. It highlights the usability of a 3D approach introduced in this paper. Results also emphasize the importance of tongue movement analysis.

Radiomics Analysis of Apparent Diffusion Coefficient Maps of Parotid Gland to Diagnose Morphologically Normal Sjogren Syndrome

OBJECTIVE

This study investigated novel radiomic features derived from apparent diffusion coefficient (ADC) maps for diagnosing Sjögren syndrome (SS) in patients without visible magnetic resonance morphologic changes.

MATERIALS AND METHODS

This study prospectively analyzed 119 consecutive patients with SS and 95 healthy volunteers using 3.0 T magnetic resonance imaging, including diffusion-weighted imaging with b values of 0 and 1000 s/mm2. Regions of interest (ROIs) were manually delineated along the margins of the largest parotid gland slice on ADC maps, from which 838 quantitative features were automatically extracted. Based on the intraclass correlation coefficient and absolute correlation coefficient, 45 radiomic parameters were selected for analysis. The differentiation between patients with SS and healthy controls was evaluated using univariate analysis and receiver operating characteristic analysis. Multiple radiomic features were integrated using binary logistic regression analysis. Through machine learning algorithms, 4 predictive models were developed, and each was thoroughly evaluated for predictive performance. The Shapley Additive exPlanations (SHAP) approach was employed to elucidate the predictive factors influencing the model.

RESULTS

Twenty-two radiomic parameters demonstrated significant differences between SS and control groups. The AUCs were 0.681 ± 0.100 (0.559~0.878). The optimal diagnostic combination for SS consisted of 6 parameters: 0.975Quantile, 180dr_D(4)_Cluster Prominence, 225dr_D(7)_Entropy, 315dr_D(7)_Entropy, Compactness2, and Max3D Diameter, achieving an AUC of 0.956. The SVM, GBM, and XGBoost models were effectively distinguished SS from healthy controls. Among all the parameters, Max3DDiameter demonstrated the strongest predictive power in the model.

CONCLUSIONS

Radiomic features derived from ADC maps demonstrate significant potential in facilitating the early diagnosis of SS.

Machine learning-driven Heckmatt grading in facioscapulohumeral muscular dystrophy: A novel pathway for musculoskeletal ultrasound analysis

OBJECTIVE

This study introduces a machine learning approach to automate muscle ultrasound analysis, aiming to improve objectivity and efficiency in segmentation, classification, and Heckmatt grading.

METHODS

We analyzed a dataset of 25,005 B-mode images from 290 participants (110 FSHD patients) acquired using a single Esaote ultrasound scanner with a standardized protocol. Manual segmentation and Heckmatt grading by experienced observers served as ground truth. K-Net was utilized for simultaneous muscle segmentation and classification. Heckmatt scoring was approached with texture analysis, using a modified scale with three classes (Normal, Uncertain, Abnormal). Radiomics features were extracted using PyRadiomics and automatic scoring was performed using XGBoost, incorporating explainability through SHAP analysis.

RESULTS

K-Net demonstrated high accuracy in skeletal muscle classification and segmentation, with Intersection over Union ranging from 73.40 to 74.03 across folds. Heckmatt's grading achieved an Area Under Curve of 0.95, 0.87, and 0.97 for classes Normal, Uncertain, and Abnormal. SHAP analysis highlighted histogram-based features as critical for visual scoring.

CONCLUSION

This study proposes and validates an automatic pipeline for muscle ultrasound analysis, leveraging machine learning for segmentation, classification, and quantitative Heckmatt grading.

SIGNIFICANCE

Automating the visual assessment of muscle ultrasound images improves the objectivity and efficiency of muscle ultrasound, supporting clinical decision-making.

Prediction of microsatellite-stable/epithelial-to-mesenchymal transition molecular subtype gastric cancer using CT radiomics and clinicopathologic factors

OBJECTIVES

This study aimed to develop a predictive model for the microsatellite-stable (MSS)/epithelial-to-mesenchymal transition (EMT) subtype of gastric cancer (GC) using computed tomography (CT) radiomics and clinicopathological factors.

MATERIALS AND METHODS

This retrospective study included 418 patients with GC who underwent primary resection and transcriptome analysis with microarray between October 1995 and May 2008. Using preoperative CT images, radiomic features from the volume of interest in the portal venous phase images were extracted. The patient data were randomly divided into training (70%) and testing (30%) datasets. Optimal radiomics features were selected through a thorough feature-selection process. The final radiomic and clinicopathological factors were selected using a stepwise variable selection method. The area under the curve (AUC) was calculated to evaluate performance.

RESULTS

Seventy patients had EMT subtype GC, and 348 patients had non-EMT subtype based on transcriptome analysis. There were 276 men (66.0 %), with a median age of 59 years (interquartile range: 50-67). Eleven radiomic features were selected for the prediction model using the combined variance inflation factor (VIF) and least absolute shrinkage and selection operator (LASSO) method. A CT radiomics-based prediction model was constructed using logistic regression with AUCs of 0.824 and 0.736 for training and testing, respectively. When clinicopathological factors such as age, tumor size, signet ring cell histology, and Lauren classification were combined, the AUCs of the models increased to 0.849 and 0.840 for training and testing, respectively (p < 0.001 for testing).

CONCLUSION

A prediction model using CT radiomics and clinicopathological factors showed good performance in predicting the EMT subtype of GC.

A comparative study of statistical, radiomics, and deep learning feature extraction techniques for medical image classification in optical and radiological modalities

Feature extraction in ML plays a crucial role in transforming raw data into a more meaningful and interpretable representation. In this study, we thoroughly examined a range of feature extraction techniques and assessed their impact on the binary classification models for medical images, utilizing a diverse and rich set of medical imaging modalities. Using H&E-stained, chest X-ray, and retina OCT images, we applied methods to extract statistical, radiomics, and deep features. These features were then used to develop PCA-LDA models as the employed classifier. We evaluated the models based on two decisive metrics: latency and performance. Latency measured the time taken for feature extraction and prediction, while mean sensitivity (balanced accuracy) characterizes the model performance. Our comparative study revealed that statistical and radiomics features were less effective for medical image classification, as they showed high latency and lower performance scores. In contrast, pre-trained DL networks performed efficiently, with high sensitivity and low latency. For H&E-stained images, the statistical feature extraction took about an hour and achieved 90.8 % sensitivity, while ResNet50 reduced processing time fourfold and increased sensitivity to 96.9 %. For chest X-rays, radiomics features were time-intensive with 92.2 % sensitivity, while ResNet50 improved sensitivity to 96 % with faster extraction time. For retina OCT images, radiomics yielded a sensitivity of 91 %, while DenseNet121 achieved 98.6 % sensitivity in 15 min. These findings underscore the superior performance of DL techniques over the statistical and radiomics features, highlighting their potential for real-world applications where accurate and rapid diagnostic decisions are essential.

Correlation of T2-Weighted Magnetic Resonance Imaging (MRI), Computed Tomography (CT), and Cone Beam Computed Tomography (CBCT) Radiomic Features for Prostate Cancer

Radiomics extracted from cone beam computed tomography (CBCT) can be assessed at time points during treatment and may provide an advantage over assessments in a pre-treatment setting using diagnostic images, like magnetic resonance imaging (MRI) or computed tomography (CT), for prostate cancer (pCa) patients receiving radiotherapy (RT). The purpose of this study was to analyze correlations between prostate radiomic features (RFs) derived from T2-weighted (T2w) MRI, CT, and first fraction CBCT for patients receiving RT for pCa. Forty-seven patients were analyzed. The prostate volumes were manually segmented, and 42 radiomic features were extracted, of which seven volume-normalized RFs were considered. The absolute Spearman correlation was calculated among the RFs of the aforementioned imaging modalities (RM) and prostate volume (RV) since the motivation of this paper was to analyze the strength of the correlation. The Benjamini-Hochberg adjustment was applied to p-values to account for multiple comparisons. No high correlations were found between CT/CBCT vs. T2w. The intramodality RM demonstrated that CT RFs were much higher than the other modalities. For example, intramodality RM≥0.95 the percentage of RFs was 17% for CT, 9% for CBCT, and 4.5% for T2w. The differences in RFs across different modalities can be viewed positively: the lack of correlation between RFs across T2w and CT/CBCT could indicate a fundamental difference in the extractable image information. It could also indicate that some RFs did not have any extractable information. A future study will include evaluating the predictive performance of patient outcomes using radiomic features from CT, CBCT, and T2w, which could help in answering such questions.

Identifying texture features from structural magnetic resonance imaging scans associated with Tourette's syndrome using machine learning

Purpose

Tourette syndrome (TS) is a neurodevelopmental disorder characterized by neurophysiological and neuroanatomical changes, primarily affecting individuals aged 2 to 18. Involuntary motor and vocal tics are common features of this syndrome. Currently, there is no curative therapy for TS, only psychological treatments or medications that temporarily manage the tics. The absence of a definitive diagnostic tool complicates the differentiation of TS from other neurological and psychological conditions.

Approach

We aim to enhance the diagnosis of TS through the classification of structural magnetic resonance scans. Our methodology comprises four sequential steps: (1) image acquisition, data were generated for the National Taiwan University, composing images of pediatric magnetic resonance imaging (MRI); (2) pre-processing, involving anatomical structural segmentation using reesurfer software; (3) feature extraction, where texture features in volumetric images are obtained; and (4) image classification, employing support vector machine and naive Bayes classifiers to determine the presence of TS.

Results

The analysis indicated significant changes in the regions of the limbic system, such as the thalamus and amygdala, and regions outside the limbic system such as medial orbitofrontal cortex and insula, which are strongly associated with TS.

Conclusions

Our findings suggest that texture features derived from sMRI scans can aid in the diagnosis of TS by highlighting critical brain regions involved in the disorder. The proposed method holds promise for improving diagnostic accuracy and understanding the neuroanatomical underpinnings of TS.

Overall Staging Prediction for Non-Small Cell Lung Cancer (NSCLC): A Local Pilot Study with Artificial Neural Network Approach

BACKGROUND

Non-small cell lung cancer (NSCLC) has been the most common cancer globally in the recent decade. CT is the most common imaging modality for the initial diagnosis of NSCLC. The gold standard for definitive diagnosis is the histological evaluation of a biopsy or surgical sample, which usually requires a long processing time for the confirmation of diagnosis. This study aims to develop artificial intelligence models to predict overall staging based on patient demographics and radiomics retrieved from the initial CT images, so as to prioritize later-stage patients for histology evaluation to facilitate cancer diagnosis.

METHOD

Two cohorts of NSCLC patient datasets were utilized for this study. The NSCLC-radiomics dataset from The Cancer Imaging Archive (TCIA) was divided into 70% for the training group and 30% for the internal testing group. Another cohort from a local hospital was collected for the an external testing group. Patient demographics and 107 radiomic features were retrieved from the gross tumor volume delineated by clinical oncologists on CT images. Artificial neural networks were used to build models for NSCLC overall staging (stage I, II, or III) prediction. Four traditional classifiers were also adopted to build models for comparison.

RESULT

The proposed feed-forward neural network (FFNN) model showed good performance in predicting overall staging with an accuracy of 88.84%, 76.67%, and 74.52% in overall accuracies in validation, internal cohort testing, and external cohort testing, respectively. The sensitivity and specificity are balanced in all the stages, with average precision and F1 score in each of the stages.

CONCLUSION

The FFNN demonstrated good performance in overall staging prediction for NSCLC patients. It has the benefit of predicting multiple overall stages in a single model. The software required and the proposed model are simple. It can be operated on a general-purpose computer in the radiology department. The application will eventually be used as a prediction tool to prioritize the biopsy or surgery sample for histological analysis and molecular investigation, thus shortening the time for diagnosis by pathologists, which supports the triage of patients for further testing.

Machine learning based classification of spontaneous intracranial hemorrhages using radiomics features

PURPOSE

To assess the efficacy of radiomics features extracted from non-contrast computed tomography (NCCT) scans in differentiating multiple etiologies of spontaneous intracerebral hemorrhage (ICH).

METHODS

CT images and clinical data from 141 ICH patients from 2010 to 2022 were collected. The cohort comprised primary (n = 57), tumorous (n = 46), and vascular malformation-related ICH (n = 38). Radiomics features were extracted from the initial brain NCCT scans and identified potential features using mutual information. A hierarchical classification with AdaBoost classifiers was employed to classify the multiple etiologies of ICH. Age of the patient and ICH's location were examined alongside radiomics features. The accuracy, area under the curve (AUC), sensitivity, and specificity were used to evaluate classification performance.

RESULTS

The proposed method achieved an accuracy of 0.79. For identifying primary ICH, the model achieved a sensitivity of 0.86 and specificity of 0.87. Meanwhile, the sensitivity and specificity for identifying tumoral causes were 0.78 and 0.93, respectively. For vascular malformation, the model reached a sensitivity and specificity of 0.72 and 0.89, respectively. The AUCs for primary, tumorous, and vascular malformation were 0.86, 0.85, and 0.82, respectively. The findings further highlight the importance of texture-based variables in ICH classification. The age and location of the ICH can enhance the classification performance.

CONCLUSION

The use of a machine learning model with radiomics features has the potential in classifying the three types of non-traumatic ICH. It may help the radiologist decide on an appropriate further examination plan to arrive at a correct diagnosis.

Identification of CT based radiomic biomarkers for progression free survival in head and neck squamous cell carcinoma

This study addresses the limited noninvasive tools for Head and Neck Squamous Cell Carcinoma (HNSCC) progression-free survival (PFS) prediction by identifying Computed Tomography (CT)-based biomarkers for predicting prognosis. A retrospective analysis was conducted on data from 203 HNSCC patients. An ensemble feature selection involving correlation analysis, univariate survival analysis, best-subset selection, and the LASSO-Cox algorithm was used to select functional features, which were then used to build final Cox Proportional Hazards models (CPH). Our CPH achieved a 0.69 concordance index in an external indepedent cohort of 77 patients. The model identified five CT-based radiomics features, Gradient ngtdm Contrast, Logσ=33D-FirstorderRootMeanSquared, Logσ=0.13D-glszm SmallAreaLowGrayLevelEmphasis, Exponential-gldm LargeDependenceHighGrayLevelEmphasis, and Gradient ngtdm Strength as survival biomarkers (p-value < 0.05). These findings contribute to our knowledge of how radiomics can be used to predict the outcome so that treatment plans can be tailored for people with HNSCC to improve their prognosis.

Radiomic analysis of cohort-specific diagnostic errors in reading dense mammograms using artificial intelligence

OBJECTIVES

This study aims to investigate radiologists' interpretation errors when reading dense screening mammograms using a radiomics-based artificial intelligence approach.

METHODS

Thirty-six radiologists from China and Australia read 60 dense mammograms. For each cohort, we identified normal areas that looked suspicious of cancer and the malignant areas containing cancers. Then radiomic features were extracted from these identified areas and random forest models were trained to recognize the areas that were most frequently linked to diagnostic errors within each cohort. The performance of the model and discriminatory power of significant radiomic features were assessed.

RESULTS

We found that in the Chinese cohort, the AUC values for predicting false positives were 0.864 (CC) and 0.829 (MLO), while in the Australian cohort, they were 0.652 (CC) and 0.747 (MLO). For false negatives, the AUC values in the Chinese cohort were 0.677 (CC) and 0.673 (MLO), and in the Australian cohort, they were 0.600 (CC) and 0.505 (MLO). In both cohorts, regions with higher Gabor and maximum response filter outputs were more prone to false positives, while areas with significant intensity changes and coarse textures were more likely to yield false negatives.

CONCLUSIONS

This cohort-based pipeline proves effective in identifying common errors for specific reader cohorts based on image-derived radiomic features.

ADVANCES IN KNOWLEDGE

This study demonstrates that radiomics-based AI can effectively identify and predict radiologists' interpretation errors in dense mammograms, with distinct radiomic features linked to false positives and false negatives in Chinese and Australian cohorts.

The value of 18F-fluorodeoxyglucose positron emission tomography-based radiomics in non-small cell lung cancer

Currently, the second most commonly diagnosed cancer in the world is lung cancer, and 85% of cases are non-small cell lung cancer (NSCLC). With growing knowledge of oncogene drivers and cancer immunology, several novel therapeutics have emerged to improve the prognostic outcomes of NSCLC. However, treatment outcomes remain diverse, and an accurate tool to achieve precision medicine is an unmet need. Radiomics, a method of extracting medical imaging features, is promising for precision medicine. Among all radiomic tools, 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET)-based radiomics provides distinct information on glycolytic activity and heterogeneity. In this review, we collected relevant literature from PubMed and summarized the various applications of 18F-FDG PET-derived radiomics in improving the detection of metastasis, subtyping histopathologies, characterizing driver mutations, assessing treatment response, and evaluating survival outcomes of NSCLC. Furthermore, we reviewed the values of 18F-FDG PET-based deep learning. Finally, several challenges and caveats exist in the implementation of 18F-FDG PET-based radiomics for NSCLC. Implementing 18F-FDG PET-based radiomics in clinical practice is necessary to ensure reproducibility. Moreover, basic studies elucidating the underlying biological significance of 18F-FDG PET-based radiomics are lacking. Current inadequacies hamper immediate clinical adoption; however, radiomic studies are progressively addressing these issues. 18F-FDG PET-based radiomics remains an invaluable and indispensable aspect of precision medicine for NSCLC.

Quantitative CT Imaging Features Associated with Stable PRISm using Machine Learning

RATIONALE AND OBJECTIVES

The structural lung features that characterize individuals with preserved ratio impaired spirometry (PRISm) that remain stable overtime are unknown. The objective of this study was to use machine learning models with computed tomography (CT) imaging to classify stable PRISm from stable controls and stable COPD and identify discriminative features.

MATERIALS AND METHODS

A total of 596 participants that did not transition between control, PRISm and COPD groups at baseline and 3-year follow-up were evaluated: n = 274 with normal lung function (stable control), n = 22 stable PRISm, and n = 300 stable COPD. Investigated features included: quantitative CT (QCT) features (n = 34), such as total lung volume (%TLCCT) and percentage of ground glass and reticulation (%GG+Reticulationtexture), as well as Radiomic (n = 102) features, including varied intensity zone distribution grainy texture (GLDZMZDV). Logistic regression machine learning models were trained using various feature combinations (Base, Base+QCT, Base+Radiomic, Base+QCT+Radiomic). Model performances were evaluated using area under receiver operator curve (AUC) and comparisons between models were made using DeLong test; feature importance was ranked using Shapley Additive Explanations values.

RESULTS

Machine learning models for all feature combinations achieved AUCs between 0.63-0.84 for stable PRISm vs. stable control, and 0.65-0.92 for stable PRISm vs. stable COPD classification. Models incorporating imaging features outperformed those trained solely on base features (p < 0.05). Compared to stable control and COPD, those with stable PRISm exhibited decreased %TLCCT and increased %GG+Reticulationtexture and GLDZMZDV.

CONCLUSION

These findings suggest that reduced lung volumes, and elevated high-density and ground glass/reticulation patterns on CT imaging are associated with stable PRISm.

Characterizing Muscle Tissue Quality Post-Stroke: Echovariation as a Clinical Indicator

Background/Objectives: Strokes remain a major global health concern, contributing significantly to disability and healthcare costs. Currently, there are no established indicators to accurately assess the degree of muscle tissue impairment in stroke-affected individuals. However, ultrasound imaging with an echotexture analysis shows potential as a quantitative tool to assess muscle tissue quality. This study aimed to identify specific echotexture features in the gastrocnemius medialis that effectively characterize muscle impairment in post-stroke individuals. Methods: An observational study was conducted with 22 post-stroke individuals. A total of 21 echotexture features were extracted and analyzed, including first-order metrics, a grey-level co-occurrence matrix, and a grey-level run length matrix. The modified Heckmatt scale was also applied to correlate with the most informative echotexture features. Results: Among the features analyzed, echovariation (EV), echointensity, and kurtosis emerged as the most informative indicators of muscle tissue quality. The EV was highlighted as the primary feature due to its strong and significant correlation with the modified Heckmatt scale (r = -0.81, p < 0.001) and its clinical and technical robustness. Lower EV values were associated with poorer muscle tissue quality, while higher values indicated better quality. Conclusions: The EV may be used as a quantitative indicator for characterizing the gastrocnemius medialis muscle tissue quality in post-stroke individuals, offering a more nuanced assessment than traditional qualitative scales. Future studies should investigate the correlation between the EV and other clinical outcomes and explore its potential to monitor the treatment efficacy, enhancing its applicability in clinical practice.

A time-dependent explainable radiomic analysis from the multi-omic cohort of CPTAC-Pancreatic Ductal Adenocarcinoma

BACKGROUND AND OBJECTIVE

In Pancreatic Ductal Adenocarcinoma (PDA), multi-omic models are emerging to answer unmet clinical needs to derive novel quantitative prognostic factors. We realized a pipeline that relies on survival machine-learning (SML) classifiers and explainability based on patients' follow-up (FU) to stratify prognosis from the public-available multi-omic datasets of the CPTAC-PDA project.

MATERIALS AND METHODS

Analyzed datasets included tumor-annotated radiologic images, clinical, and mutational data. A feature selection was based on univariate (UV) and multivariate (MV) survival analyses according to Overall Survival (OS) and recurrence (REC). In this study, we considered seven multi-omic datasets and compared four SML classifiers: Cox, survival random forest, generalized boosted, and support vector machines (SVM). For each classifier, we assessed the concordance (C) index on the validation set. The best classifiers for the validation set on both OS and REC underwent explainability analyses using SurvSHAP(t), which extends SHapley Additive exPlanations (SHAP).

RESULTS

According to OS, after UV and MV analyses we selected 18/37 and 10/37 multi-omic features, respectively. According to REC, based on UV and MV analyses we selected 10/35 and 5/35 determinants, respectively. Generally, SML classifiers including radiomics outperformed those modelled on clinical or mutational predictors. For OS, the Cox model encompassing radiomic, clinical, and mutational features reached 75 % of C index, outperforming other classifiers. On the other hand, for REC, the SVM model including only radiomics emerged as the best-performing, with 68 % of C index. For OS, SurvSHAP(t) identified the first order Median Gray Level (GL) intensities, the gender, the tumor grade, the Joint Energy GL Co-occurrence Matrix (GLCM), and the GLCM Informational Measures of Correlations of type 1 as the most important features. For REC, the first order Median GL intensities, the GL size zone matrix Small Area Low GL Emphasis, and first order variance of GL intensities emerged as the most discriminative.

CONCLUSIONS

In this work, radiomics showed the potential for improving patients' risk stratification in PDA. Furthermore, a deeper understanding of how radiomics can contribute to prognosis in PDA was achieved with a time-dependent explainability of the top multi-omic predictors.

Image biomarkers and explainable AI: handcrafted features versus deep learned features

Feature extraction and selection from medical data are the basis of radiomics and image biomarker discovery for various architectures, including convolutional neural networks (CNNs). We herein describe the typical radiomics steps and the components of a CNN for both deep feature extraction and end-to-end approaches. We discuss the curse of dimensionality, along with dimensionality reduction techniques. Despite the outstanding performance of deep learning (DL) approaches, the use of handcrafted features instead of deep learned features needs to be considered for each specific study. Dataset size is a key factor: large-scale datasets with low sample diversity could lead to overfitting; limited sample sizes can provide unstable models. The dataset must be representative of all the "facets" of the clinical phenomenon/disease investigated. The access to high-performance computational resources from graphics processing units is another key factor, especially for the training phase of deep architectures. The advantages of multi-institutional federated/collaborative learning are described. When large language models are used, high stability is needed to avoid catastrophic forgetting in complex domain-specific tasks. We highlight that non-DL approaches provide model explainability superior to that provided by DL approaches. To implement explainability, the need for explainable AI arises, also through post hoc mechanisms. RELEVANCE STATEMENT: This work aims to provide the key concepts for processing the imaging features to extract reliable and robust image biomarkers. KEY POINTS: The key concepts for processing the imaging features to extract reliable and robust image biomarkers are provided. The main differences between radiomics and representation learning approaches are highlighted. The advantages and disadvantages of handcrafted versus learned features are given without losing sight of the clinical purpose of artificial intelligence models.

Ensemble feature selection and tabular data augmentation with generative adversarial networks to enhance cutaneous melanoma identification and interpretability

BACKGROUND

Cutaneous melanoma is the most aggressive form of skin cancer, responsible for most skin cancer-related deaths. Recent advances in artificial intelligence, jointly with the availability of public dermoscopy image datasets, have allowed to assist dermatologists in melanoma identification. While image feature extraction holds potential for melanoma detection, it often leads to high-dimensional data. Furthermore, most image datasets present the class imbalance problem, where a few classes have numerous samples, whereas others are under-represented.

METHODS

In this paper, we propose to combine ensemble feature selection (FS) methods and data augmentation with the conditional tabular generative adversarial networks (CTGAN) to enhance melanoma identification in imbalanced datasets. We employed dermoscopy images from two public datasets, PH2 and Derm7pt, which contain melanoma and not-melanoma lesions. To capture intrinsic information from skin lesions, we conduct two feature extraction (FE) approaches, including handcrafted and embedding features. For the former, color, geometric and first-, second-, and higher-order texture features were extracted, whereas for the latter, embeddings were obtained using ResNet-based models. To alleviate the high-dimensionality in the FE, ensemble FS with filter methods were used and evaluated. For data augmentation, we conducted a progressive analysis of the imbalance ratio (IR), related to the amount of synthetic samples created, and evaluated the impact on the predictive results. To gain interpretability on predictive models, we used SHAP, bootstrap resampling statistical tests and UMAP visualizations.

RESULTS

The combination of ensemble FS, CTGAN, and linear models achieved the best predictive results, achieving AUCROC values of 87% (with support vector machine and IR=0.9) and 76% (with LASSO and IR=1.0) for the PH2 and Derm7pt, respectively. We also identified that melanoma lesions were mainly characterized by features related to color, while not-melanoma lesions were characterized by texture features.

CONCLUSIONS

Our results demonstrate the effectiveness of ensemble FS and synthetic data in the development of models that accurately identify melanoma. This research advances skin lesion analysis, contributing to both melanoma detection and the interpretation of main features for its identification.

Predicting regional tau accumulation with machine learning-based tau-PET and advanced radiomics

INTRODUCTION

Alzheimer's disease is partially characterized by the progressive accumulation of aggregated tau-containing neurofibrillary tangles. Although the association between accumulated tau, neurodegeneration, and cognitive decline is critical for disease understanding and clinical trial design, we still lack robust tools to predict individualized trajectories of tau accumulation. Our objective was to assess whether brain imaging biomarkers of flortaucipir-positron emission tomography (PET), in combination with clinical and genomic measures, could predict future pathological tau accumulation.

METHODS

We quantified the disease profile of participants (N = 276) using a comprehensive set of descriptors, including clinical/demographic (age, diagnosis, amyloid status, sex, race, ethnicity), genetic (apolipoprotein E [APOE]-ε4), and flortaucipir-PET imaging measures (regional flortaucipir standardized uptake value ratio [SUVr] and comprehensive radiomic texture features extracted from Automated Anatomical Labeling template regions). We trained an AdaBoost machine learning algorithm in a 2:1 split train-test configuration to derive a prognostic index that (i) stratifies individualized brain regions including global (AD-signature region) and lobar regions (frontal, occipital, parietal, temporal) into stable/slow- and fast-progressors based on future tau accumulation, and (ii) forecasts individualized regional annualized-rate-of-change in flortaucipir-PET SUVr. Further, we developed an adaptive model incorporating flortaucipir-PET measurements from the baseline and intermediate timepoints to predict annualized-rate-of-change.

RESULTS

In binary classification for predicting stable/slow- versus fast-progressors, the area-under-the-receiver-operating-characteristic curve was 0.86 in the AD-signature region and 0.83, 0.82, 0.84, and 0.83 in frontal, occipital, parietal, and temporal regions, respectively. The trained models successfully predicted annualized-rate-of-change of flortaucipir-PET regional flortaucipir SUVr in AD-signature and lobar regions (Pearson-correlation [R]: AD-signature = 0.73; frontal = 0.73; occipital = 0.71; parietal = 0.70; temporal = 0.69). The models' performance in predicting annualized-rate-of-change slightly increased when imaging features from intermediate timepoints were used in the adaptive setting (R: AD-signature = 0.79; frontal = 0.87; occipital = 0.83; parietal = 0.74; temporal = 0.82).

DISCUSSION

Taken together, our results propose a robust approach to predict future tau accumulation that may improve the ability to enroll, stratify, and gauge efficacy in clinical trial participants.

Highlights

Machine learning predicts the future rate of tau accumulation in Alzheimer's disease.Tau prediction in lobar/global regions benefits from diverse multimodal features.This prognostic index can serve as a sensitive tool for patient stratification.

Local Assessment and Small Bowel Crohn's Disease Severity Scoring using AI

RATIONALE AND OBJECTIVES

We present a machine learning and computer vision approach for a localized, automated, and standardized scoring of Crohn's disease (CD) severity in the small bowel, overcoming the current limitations of manual measurements CT enterography (CTE) imaging and qualitative assessments, while also considering the complex anatomy and distribution of the disease.

MATERIALS AND METHODS

Two radiologists introduced a severity score and evaluated disease severity at 7.5 mm intervals along the curved planar reconstruction of the distal and terminal ileum using 236 CTE scans. A hybrid model, combining deep-learning, 3-D CNN, and Random Forest model, was developed to classify disease severity at each mini-segment. Precision, sensitivity, weighted Cohen's score, and accuracy were evaluated on a 20% hold-out test set.

RESULTS

The hybrid model achieved precision and sensitivity ranging from 42.4% to 84.1% for various severity categories (normal, mild, moderate, and severe) on the test set. The model's Cohen's score (κ = 0.83) and accuracy (70.7%) were comparable to the inter-observer agreement between experienced radiologists (κ = 0.87, accuracy = 76.3%). The model accurately predicted disease length, correlated with radiologist-reported disease length (r = 0.83), and accurately identified the portion of total ileum containing moderate-to-severe disease with an accuracy of 91.51%.

CONCLUSION

The proposed automated hybrid model offers a standardized, reproducible, and quantitative local assessment of small bowel CD severity and demonstrates its value in CD severity assessment.

T2-FLAIR imaging-based radiomic features for predicting early postoperative recurrence of grade II gliomas

Aim: To develop and validate a T2-weighted-fluid attenuated inversion recovery (T2-FLAIR) images-based radiomics model for predicting early postoperative recurrence (within 1 year) in patients with low-grade gliomas (LGGs).Methods: A retrospective analysis was performed by collecting clinical, pathological and magnetic resonance imaging (MRI) data from patients with LGG between 2017 and 2022. Regions of interest were delineated and radiomic features were extracted from T2-FLAIR images using 3D-Slicer software. To minimize redundant features, the Least Absolute Shrinkage and Selection Operator (LASSO) regression algorithm was used. Patients were categorized into two groups based on recurrence status: the recurrence group (RG) and the non-recurrence group (NRG). Radiomic features were used to develop models using three machine learning approaches: logistic regression (LR), random forest (RF) and support vector machine (SVM). The performance of the radiomic features was validated using fivefold cross-validation.Results: After rigorous screening, 105 patients met the inclusion criteria, and five radiomic features were identified. After 5-folds cross-validation, the average areas under the curves for LR, RF and SVM were 0.813, 0.741 and 0.772, respectively.Conclusion: T2-FLAIR-based radiomic features effectively predicted early recurrence in postoperative LGGs.

Quantifying the reproducibility and longitudinal repeatability of radiomics features in magnetic resonance Image-Guide accelerator Imaging: A phantom study

OBJECTIVE

This study aimed to quantitatively evaluate the inter-platform reproducibility and longitudinal acquisition repeatability of MRI radiomics features in Fluid-Attenuated Inversion Recovery (FLAIR), T2-weighted (T2W), and T1-weighted (T1W) sequences on MR-Linac systems using an American College of Radiology (ACR) phantom.

MATERIALS AND METHODS

This study used two MR-Linac systems (A and B) in different cancer centers. The ACR phantom was scanned on system A daily for 30 consecutive days to evaluate longitudinal repeatability. Additionally, retest data were collected after repositioning the phantom. Inter-platform reproducibility was assessed by conducting scans under identical conditions using system B. Regions of interest were delineated on the T1W sequence from system A and mapped to other sequences via rigid registration. Intra-observer and inter-observer comparisons were conducted. Repeatability and reproducibility were assessed using the intraclass correlation coefficient (ICC) and coefficient of variation (CV). Robust radiomics features were identified based on ICC>0.9 and CV<10 %.

RESULTS

Analysis showed that a higher proportion of radiomics features derived from longitudinal FLAIR sequence (51.65 %) met robustness criteria compared to T2W (48.35 %) and T1W (43.96 %). Additionally, more inter-platform features from the FLAIR sequence (62.64 %) were robust compared to T2W (42.86 %) and T1W (39.56 %). Test-retest and intra-observer repeatability were excellent across all sequences, with a median ICC of 0.99 and CV<5%. However, inter-observer reproducibility was inferior, especially for the T1W sequence.

CONCLUSIONS

Different sequences show variations in repeatability and reproducibility. The FLAIR sequence demonstrated advantages in both longitudinal repeatability and inter-platform reproducibility. Caution is warranted when interpreting data, particularly in longitudinal or multiplatform radiomics studies.

Glioma Tumor Grading Using Radiomics on Conventional MRI: A Comparative Study of WHO 2021 and WHO 2016 Classification of Central Nervous Tumors

BACKGROUND

Glioma grading transformed in World Health Organization (WHO) 2021 CNS tumor classification, integrating molecular markers. However, the impact of this change on radiomics-based machine learning (ML) classifiers remains unexplored.

PURPOSE

To assess the performance of ML in classifying glioma tumor grades based on various WHO criteria.

STUDY TYPE

Retrospective.

SUBJECTS

A neuropathologist regraded gliomas of 237 patients into WHO 2016 and 2021 from 2007 criteria.

FIELD STRENGTH/SEQUENCE

Multicentric 0.5 to 3 Tesla; pre- and post-contrast T1-weighted, T2-weighted, and fluid-attenuated inversion recovery.

ASSESSMENT

Radiomic features were selected using random forest-recursive feature elimination. The synthetic minority over-sampling technique (SMOTE) was implemented for data augmentation. Stratified 10-fold cross-validation with and without SMOTE was used to evaluate 11 classifiers for 3-grade (2, 3, and 4; WHO 2016 and 2021) and 2-grade (low and high grade; WHO 2007 and 2021) classification. Additionally, we developed the models on data randomly divided into training and test sets (mixed-data analysis), or data divided based on the centers (independent-data analysis).

STATISTICAL TESTS

We assessed ML classifiers using sensitivity, specificity, accuracy, and the area under the receiver operating characteristic curve (AUC). Top performances were compared with a t-test and categorical data with the chi-square test using a significance level of P < 0.05.

RESULTS

In the mixed-data analysis, Stacking Classifier without SMOTE achieved the highest accuracy (0.86) and AUC (0.92) in 3-grade WHO 2021 grouping. The results of WHO 2021 were significantly better than WHO 2016 (P-value<0.0001). In the 2-grade analysis, ML achieved 1.00 in all metrics. In the independent-data analysis, ML classifiers showed strong discrimination between grade 2 and 4, despite lower performance metrics than the mixed analysis.

DATA CONCLUSION

ML algorithms performed better in glioma tumor grading based on WHO 2021 criteria. Nonetheless, the clinical use of ML classifiers needs further investigation.

LEVEL OF EVIDENCE

3 TECHNICAL EFFICACY: Stage 2.

Identification and Quantification of Precursory Changes of Rheumatoid Vasculitis in the Dorsalis Pedis Artery

OBJECTIVE

Rheumatoid arthritis (RA) is a systemic connective tissue autoimmune disease that can infiltrate arterial walls. The delay in diagnosis and treatment of rheumatoid vasculitis (RV) in patients with RA may lead to irreversible damage to the arterial walls of small-to-medium vessels, which has serious and devastating consequences, most notably lung and cardiac damage. In this work an ultrasound image-based biomarker was developed to detect precursory changes in RV.

METHODS

The ground truth was initiated from a medical diagnosis of RA, with arterial wall thickening of the proximal dorsalis pedis artery (DPA) indicating precursory changes of RV identified with ultrasound scanning. Ultrasound images of the DPA from 49 healthy subjects in the control group and 46 patients in the RA group were obtained. In total, 187 texture features were extracted from the images, followed by principal component analysis and linear discriminant analysis.

RESULTS

The proposed biomarker detected a significant difference between the two groups (p = 5.74 × 10-18) with an area under the receiver operating characteristic curve of 0.85. Ten major textural features contributing most heavily to the biomarker were identified, with these textures being consistent with clinical observations of RV identified in previous studies. Interscan reproducibility was assessed by computing the biomarker twice based on repeated scans of each ankle. High interscan reproducibility was demonstrated by a strong and significant Pearson's coefficient (r = 0.85, p < 0.01) between the two repeated measurements of the proposed biomarker.

CONCLUSION

The proposed biomarker can discriminate image textural differences seen in images acquired from RA patients, demonstrating precursory changes in RV compared with healthy controls. The major discriminative features identified in this study may facilitate the early identification and treatment of RV.

CT radiomics-based biomarkers can predict response to immunotherapy in hepatocellular carcinoma

Hepatocellular Carcinoma (HCC) remains a leading cause of cancer deaths. Despite the rise of immunotherapies, many HCC patients don't benefit. There's a clear need for biomarkers to guide treatment decisions. This research aims to identify such biomarkers by combining radiological data and machine learning. We analyzed clinical and CT imaging data of 54 HCC patients undergoing immunotherapy. Radiologic features were examined to develop a model predicting short-term immunotherapy effects. We utilized 9 machine learning and 2 ensemble learning techniques using RapidMiner for model construction. We conducted the validation of the above feature combination using 29 HCC patients who received immunotherapy from another hospital, and tested and validated it using XGBoost combined with a genetic algorithm. In 54 HCC patients, radiomics features varied significantly between those with partial response (PR) and stable disease (SD). Key features in Gray Level Run Length Matrix (GLRLM) and in adjacent tissues' Intensity Direct, Neighborhood Gray Tone Difference Matrix (NGTDM), and Shape correlated with short-term immunotherapy efficacy. Selected feature combinations of 15, 19, and 8/15 features yielded better outcomes. Deep learning, random forest, and naive bayes outperformed other methods, with Bagging being more reliable than Adaboost. In the validation set of 29 HCC patients, the mentioned feature combination also demonstrated favorable performance. Furthermore, we achieved improved results when employing XGBoost in conjunction with a genetic algorithm for testing and validation. The machine learning model built with CT image features derived from GLCM, GLRLM, IntensityDirect, NGTDM, and Shape can accurately forecast the short-term efficacy of immunotherapy for HCC.

Hybridization of Acoustic and Visual Features of Polish Sibilants Produced by Children for Computer Speech Diagnosis

Speech disorders are significant barriers to the balanced development of a child. Many children in Poland are affected by lisps (sigmatism)-the incorrect articulation of sibilants. Since speech therapy diagnostics is complex and multifaceted, developing computer-assisted methods is crucial. This paper presents the results of assessing the usefulness of hybrid feature vectors extracted based on multimodal (video and audio) data for the place of articulation assessment in sibilants /s/ and /ʂ/. We used acoustic features and, new in this field, visual parameters describing selected articulators' texture and shape. Analysis using statistical tests indicated the differences between various sibilant realizations in the context of the articulation pattern assessment using hybrid feature vectors. In sound /s/, 35 variables differentiated dental and interdental pronunciation, and 24 were visual (textural and shape). For sibilant /ʂ/, we found 49 statistically significant variables whose distributions differed between speaker groups (alveolar, dental, and postalveolar articulation), and the dominant feature type was noise-band acoustic. Our study suggests hybridizing the acoustic description with video processing provides richer diagnostic information.

A radiomics nomogram based on multiparametric MRI for diagnosing focal cortical dysplasia and initially identifying laterality

BACKGROUND

Focal cortical dysplasia (FCD) is the most common epileptogenic developmental malformation. The diagnosis of FCD is challenging. We generated a radiomics nomogram based on multiparametric magnetic resonance imaging (MRI) to diagnose FCD and identify laterality early.

METHODS

Forty-three patients treated between July 2017 and May 2022 with histopathologically confirmed FCD were retrospectively enrolled. The contralateral unaffected hemispheres were included as the control group. Therefore, 86 ROIs were finally included. Using January 2021 as the time cutoff, those admitted after January 2021 were included in the hold-out set (n = 20). The remaining patients were separated randomly (8:2 ratio) into training (n = 55) and validation (n = 11) sets. All preoperative and postoperative MR images, including T1-weighted (T1w), T2-weighted (T2w), fluid-attenuated inversion recovery (FLAIR), and combined (T1w + T2w + FLAIR) images, were included. The least absolute shrinkage and selection operator (LASSO) was used to select features. Multivariable logistic regression analysis was used to develop the diagnosis model. The performance of the radiomic nomogram was evaluated with an area under the curve (AUC), net reclassification improvement (NRI), integrated discrimination improvement (IDI), calibration and clinical utility.

RESULTS

The model-based radiomics features that were selected from combined sequences (T1w + T2w + FLAIR) had the highest performances in all models and showed better diagnostic performance than inexperienced radiologists in the training (AUCs: 0.847 VS. 0.664, p = 0.008), validation (AUC: 0.857 VS. 0.521, p = 0.155), and hold-out sets (AUCs: 0.828 VS. 0.571, p = 0.080). The positive values of NRI (0.402, 0.607, 0.424) and IDI (0.158, 0.264, 0.264) in the three sets indicated that the diagnostic performance of Model-Combined improved significantly. The radiomics nomogram fit well in calibration curves (p > 0.05), and decision curve analysis further confirmed the clinical usefulness of the nomogram. Additionally, the contrast (the radiomics feature) of the FCD lesions not only played a crucial role in the classifier but also had a significant correlation (r = -0.319, p < 0.05) with the duration of FCD.

CONCLUSION

The radiomics nomogram generated by logistic regression model-based multiparametric MRI represents an important advancement in FCD diagnosis and treatment.

Cardiovascular Magnetic Resonance Radiomics to Identify Components of the Extracellular Matrix in Dilated Cardiomyopathy

BACKGROUND

Current cardiovascular magnetic resonance sequences cannot discriminate between different myocardial extracellular space (ECSs), including collagen, noncollagen, and inflammation. We sought to investigate whether cardiovascular magnetic resonance radiomics analysis can distinguish between noncollagen and inflammation from collagen in dilated cardiomyopathy.

METHODS

We identified data from 132 patients with dilated cardiomyopathy scheduled for an invasive septal biopsy who underwent cardiovascular magnetic resonance at 3 T. Cardiovascular magnetic resonance imaging protocol included native and postcontrast T1 mapping and late gadolinium enhancement (LGE). Radiomic features were computed from the midseptal myocardium, near the biopsy region, on native T1, extracellular volume (ECV) map, and LGE images. Principal component analysis was used to reduce the number of radiomic features to 5 principal radiomics. Moreover, a correlation analysis was conducted to identify radiomic features exhibiting a strong correlation (r>0.9) with the 5 principal radiomics. Biopsy samples were used to quantify ECS, myocardial fibrosis, and inflammation.

RESULTS

Four histopathological phenotypes were identified: low collagen (n=20), noncollagenous ECS expansion (n=49), mild to moderate collagenous ECS expansion (n=42), and severe collagenous ECS expansion (n=21). Noncollagenous expansion was associated with the highest risk of myocardial inflammation (65%). Although native T1 and ECV provided high diagnostic performance in differentiating severe fibrosis (C statistic, 0.90 and 0.90, respectively), their performance in differentiating between noncollagen and mild to moderate collagenous expansion decreased (C statistic: 0.59 and 0.55, respectively). Integration of ECV principal radiomics provided better discrimination and reclassification between noncollagen and mild to moderate collagen (C statistic, 0.79; net reclassification index, 0.83 [95% CI, 0.45-1.22]; P<0.001). There was a similar trend in the addition of native T1 principal radiomics (C statistic, 0.75; net reclassification index, 0.93 [95% CI, 0.56-1.29]; P<0.001) and LGE principal radiomics (C statistic, 0.74; net reclassification index, 0.59 [95% CI, 0.19-0.98]; P=0.004). Five radiomic features per sequence were identified with correlation analysis. They showed a similar improvement in performance for differentiating between noncollagen and mild to moderate collagen (native T1, ECV, LGE C statistic, 0.75, 0.77, and 0.71, respectively). These improvements remained significant when confined to a single radiomic feature (native T1, ECV, LGE C statistic, 0.71, 0.70, and 0.64, respectively).

CONCLUSIONS

Radiomic features extracted from native T1, ECV, and LGE provide incremental information that improves our capability to discriminate noncollagenous expansion from mild to moderate collagen and could be useful for detecting subtle chronic inflammation in patients with dilated cardiomyopathy.

Artificial intelligence-based, semi-automated segmentation for the extraction of ultrasound-derived radiomics features in breast cancer: a prospective multicenter study

PURPOSE

To investigate the feasibility of an artificial intelligence (AI)-based semi-automated segmentation for the extraction of ultrasound (US)-derived radiomics features in the characterization of focal breast lesions (FBLs).

MATERIAL AND METHODS

Two expert radiologists classified according to US BI-RADS criteria 352 FBLs detected in 352 patients (237 at Center A and 115 at Center B). An AI-based semi-automated segmentation was used to build a machine learning (ML) model on the basis of B-mode US of 237 images (center A) and then validated on an external cohort of B-mode US images of 115 patients (Center B).

RESULTS

A total of 202 of 352 (57.4%) FBLs were benign, and 150 of 352 (42.6%) were malignant. The AI-based semi-automated segmentation achieved a success rate of 95.7% for one reviewer and 96% for the other, without significant difference (p = 0.839). A total of 15 (4.3%) and 14 (4%) of 352 semi-automated segmentations were not accepted due to posterior acoustic shadowing at B-Mode US and 13 and 10 of them corresponded to malignant lesions, respectively. In the validation cohort, the characterization made by the expert radiologist yielded values of sensitivity, specificity, PPV and NPV of 0.933, 0.9, 0.857, 0.955, respectively. The ML model obtained values of sensitivity, specificity, PPV and NPV of 0.544, 0.6, 0.416, 0.628, respectively. The combined assessment of radiologists and ML model yielded values of sensitivity, specificity, PPV and NPV of 0.756, 0.928, 0.872, 0.855, respectively.

CONCLUSION

AI-based semi-automated segmentation is feasible, allowing an instantaneous and reproducible extraction of US-derived radiomics features of FBLs. The combination of radiomics and US BI-RADS classification led to a potential decrease of unnecessary biopsy but at the expense of a not negligible increase of potentially missed cancers.

MRI radiomics enhances radiologists' ability for characterizing intestinal fibrosis in patients with Crohn's disease

OBJECTIVES

We aimed to develop MRI-based radiomic models (RMs) to improve the diagnostic accuracy of radiologists in characterizing intestinal fibrosis in patients with Crohn's disease (CD).

METHODS

This retrospective study included patients with refractory CD who underwent MR before surgery from November 2013 to September 2021. Resected bowel segments were histologically classified as none-mild or moderate-severe fibrosis. RMs based on different MR sequence combinations (RM1: T2WI and enhanced-T1WI; RM2: T2WI, enhanced-T1WI, diffusion-weighted imaging [DWI], and apparent diffusion coefficient [ADC]); RM3: T2WI, enhanced-T1WI, DWI, ADC, and magnetization transfer MRI [MTI]), were developed and validated in an independent test cohort. The RMs' diagnostic performance was compared to that of visual interpretation using identical sequences and a clinical model.

RESULTS

The final population included 123 patients (81 men, 42 women; mean age: 30.26 ± 7.98 years; training cohort, n = 93; test cohort, n = 30). The area under the receiver operating characteristic curve (AUC) of RM1, RM2, and RM3 was 0.86 (p = 0.001), 0.88 (p = 0.001), and 0.93 (p = 0.02), respectively. The decision curve analysis confirmed a progressive improvement in the diagnostic performance of three RMs with the addition of more specific sequences. All RMs performance surpassed the visual interpretation based on the same MR sequences (visual model 1, AUC = 0.65, p = 0.56; visual model 2, AUC = 0.63, p = 0.04; visual model 3, AUC = 0.77, p = 0.002), as well as the clinical model composed of C-reactive protein and erythrocyte sedimentation rate (AUC = 0.60, p = 0.13).

CONCLUSIONS

The RMs, utilizing various combinations of conventional, DWI and MTI sequences, significantly enhance radiologists' ability to accurately characterize intestinal fibrosis in patients with CD.

CRITICAL RELEVANCE STATEMENT

The utilization of MRI-based RMs significantly enhances the diagnostic accuracy of radiologists in characterizing intestinal fibrosis.

KEY POINTS

MRI-based RMs can characterize CD intestinal fibrosis using conventional, diffusion, and MTI sequences. The RMs achieved AUCs of 0.86-0.93 for assessing fibrosis grade. MRI-radiomics outperformed visual interpretation for grading CD intestinal fibrosis.

Development and Validation of Prognostic Models Using Radiomic Features from Pre-Treatment Positron Emission Tomography (PET) Images in Head and Neck Squamous Cell Carcinoma (HNSCC) Patients

High-dimensional radiomics features derived from pre-treatment positron emission tomography (PET) images offer prognostic insights for patients with head and neck squamous cell carcinoma (HNSCC). Using 124 PET radiomics features and clinical variables (age, sex, stage of cancer, site of cancer) from a cohort of 232 patients, we evaluated four survival models-penalized Cox model, random forest, gradient boosted model and support vector machine-to predict all-cause mortality (ACM), locoregional recurrence/residual disease (LR) and distant metastasis (DM) probability during 36, 24 and 24 months of follow-up, respectively. We developed models with five-fold cross-validation, selected the best-performing model for each outcome based on the concordance index (C-statistic) and the integrated Brier score (IBS) and validated them in an independent cohort of 102 patients. The penalized Cox model demonstrated better performance for ACM (C-statistic = 0.70, IBS = 0.12) and DM (C-statistic = 0.70, IBS = 0.08) while the random forest model displayed better performance for LR (C-statistic = 0.76, IBS = 0.07). We conclude that the ML-based prognostic model can aid clinicians in quantifying prognosis and determining effective treatment strategies, thereby improving favorable outcomes in HNSCC patients.

Prediction of cervix cancer stage and grade from diffusion weighted imaging using EfficientNet

Purpose. This study aims to introduce an innovative noninvasive method that leverages a single image for both grading and staging prediction. The grade and the stage of cervix cancer (CC) are determined from diffusion-weighted imaging (DWI) in particular apparent diffusion coefficient (ADC) maps using deep convolutional neural networks (DCNN).Methods. datasets composed of 85 patients having annotated tumor stage (I, II, III, and IV), out of this, 66 were with grade (II and III) and the remaining patients with no reported grade were retrospectively collected. The study was IRB approved. For each patient, sagittal and axial slices containing the gross tumor volume (GTV) were extracted from ADC maps. These were computed using the mono exponential model from diffusion weighted images (b-values = 0, 100, 1000) that were acquired prior to radiotherapy treatment. Balanced training sets were created using the Synthetic Minority Oversampling Technique (SMOTE) and fed to the DCNN. EfficientNetB0 and EfficientNetB3 were transferred from the ImageNet application to binary and four-class classification tasks. Five-fold stratified cross validation was performed for the assessment of the networks. Multiple evaluation metrics were computed including the area under the receiver operating characteristic curve (AUC). Comparisons with Resnet50, Xception, and radiomic analysis were performed.Results. for grade prediction, EfficientNetB3 gave the best performance with AUC = 0.924. For stage prediction, EfficientNetB0 was the best with AUC = 0.931. The difference between both models was, however, small and not statistically significant EfficientNetB0-B3 outperformed ResNet50 (AUC = 0.71) and Xception (AUC = 0.89) in stage prediction, and demonstrated comparable results in grade classification, where AUCs of 0.89 and 0.90 were achieved by ResNet50 and Xception, respectively. DCNN outperformed radiomic analysis that gave AUC = 0.67 (grade) and AUC = 0.66 (stage).Conclusion.the prediction of CC grade and stage from ADC maps is feasible by adapting EfficientNet approaches to the medical context.

Interpretable Radiomic Signature for Breast Microcalcification Detection and Classification

Breast microcalcifications are observed in 80% of mammograms, and a notable proportion can lead to invasive tumors. However, diagnosing microcalcifications is a highly complicated and error-prone process due to their diverse sizes, shapes, and subtle variations. In this study, we propose a radiomic signature that effectively differentiates between healthy tissue, benign microcalcifications, and malignant microcalcifications. Radiomic features were extracted from a proprietary dataset, composed of 380 healthy tissue, 136 benign, and 242 malignant microcalcifications ROIs. Subsequently, two distinct signatures were selected to differentiate between healthy tissue and microcalcifications (detection task) and between benign and malignant microcalcifications (classification task). Machine learning models, namely Support Vector Machine, Random Forest, and XGBoost, were employed as classifiers. The shared signature selected for both tasks was then used to train a multi-class model capable of simultaneously classifying healthy, benign, and malignant ROIs. A significant overlap was discovered between the detection and classification signatures. The performance of the models was highly promising, with XGBoost exhibiting an AUC-ROC of 0.830, 0.856, and 0.876 for healthy, benign, and malignant microcalcifications classification, respectively. The intrinsic interpretability of radiomic features, and the use of the Mean Score Decrease method for model introspection, enabled models' clinical validation. In fact, the most important features, namely GLCM Contrast, FO Minimum and FO Entropy, were compared and found important in other studies on breast cancer.

Robustness and classification capabilities of MRI radiomic features in identifying carotid plaque vulnerability

OBJECTIVES

To assess how radiomic features may be combined with plaque morphological and compositional features identified by multi-contrast MRI to improve upon conventional risk assessment models in determining culprit carotid artery lesions.

METHODS

Fifty-five patients (mean age: 62.6; 35 males) with bilateral carotid stenosis who experienced transient ischaemic attack (TIA) or stroke were included from the CARE-II multi-centre carotid imaging trial (ClinicalTrials.gov Identifier: NCT02017756). They underwent MRI within 2 weeks of the event. Classification capability in distinguishing culprit lesions was assessed by machine learning. Repeatability and reproducibility of the results were investigated by assessing the robustness of the radiomic features.

RESULTS

Radiomics combined with a relatively conventional plaque morphological and compositional metric-based model provided incremental value over a conventional model alone (area under curve [AUC], 0.819 ± 0.002 vs 0.689 ± 0.019, respectively, P = .014). The radiomic model alone also provided value over the conventional model (AUC, 0.805 ± 0.003 vs 0.689 ± 0.019, respectively, P = .031). T2-weighted imaging-based radiomic features had consistently higher robustness and classification capabilities compared with T1-weighted images. Higher-dimensional radiomic features outperformed first-order features. Grey Level Co-occurrence Matrix, Grey Level Dependence Matrix, and Grey Level Size Zone Matrix sub-types were particularly useful in identifying textures which could detect vulnerable lesions.

CONCLUSIONS

The combination of MRI-based radiomic features and lesion morphological and compositional parameters provided added value to the reference-standard risk assessment for carotid atherosclerosis. This may improve future risk stratification for individuals at risk of major adverse ischaemic cerebrovascular events.

ADVANCES IN KNOWLEDGE

The clinical relevance of this work is that it addresses the need for a more comprehensive method of risk assessment for patients at risk of ischaemic stroke, beyond conventional stenosis measurement. This paper shows that in the case of carotid stroke, high-dimensional radiomics features can improve classification capabilities compared with stenosis measurement alone.

Radiomic biomarkers of locoregional recurrence: prognostic insights from oral cavity squamous cell carcinoma preoperative CT scans

Introduction

This study aimed to identify CT-based imaging biomarkers for locoregional recurrence (LR) in Oral Cavity Squamous Cell Carcinoma (OSCC) patients.

Methods

Computed tomography scans were collected from 78 patients with OSCC who underwent surgical treatment at a single medical center. We extracted 1,092 radiomic features from gross tumor volume in each patient's pre-treatment CT. Clinical characteristics were also obtained, including race, sex, age, tobacco and alcohol use, tumor staging, and treatment modality. A feature selection algorithm was used to eliminate the most redundant features, followed by a selection of the best subset of the Logistic regression model (LRM). The best LRM model was determined based on the best prediction accuracy in terms of the area under Receiver operating characteristic curve. Finally, significant radiomic features in the final LRM model were identified as imaging biomarkers.

Results and discussion

Two radiomics biomarkers, Large Dependence Emphasis (LDE) of the Gray Level Dependence Matrix (GLDM) and Long Run Emphasis (LRE) of the Gray Level Run Length Matrix (GLRLM) of the 3D Laplacian of Gaussian (LoG σ=3), have demonstrated the capability to preoperatively distinguish patients with and without LR, exhibiting exceptional testing specificity (1.00) and sensitivity (0.82). The group with LRE > 2.99 showed a 3-year recurrence-free survival rate of 0.81, in contrast to 0.49 for the group with LRE ≤ 2.99. Similarly, the group with LDE > 120 showed a rate of 0.82, compared to 0.49 for the group with LDE ≤ 120. These biomarkers broaden our understanding of using radiomics to predict OSCC progression, enabling personalized treatment plans to enhance patient survival.

Developing a novel image marker to predict the clinical outcome of neoadjuvant chemotherapy (NACT) for ovarian cancer patients

OBJECTIVE

Neoadjuvant chemotherapy (NACT) is one kind of treatment for advanced stage ovarian cancer patients. However, due to the nature of tumor heterogeneity, the clinical outcomes to NACT vary significantly among different subgroups. Partial responses to NACT may lead to suboptimal debulking surgery, which will result in adverse prognosis. To address this clinical challenge, the purpose of this study is to develop a novel image marker to achieve high accuracy prognosis prediction of NACT at an early stage.

METHODS

For this purpose, we first computed a total of 1373 radiomics features to quantify the tumor characteristics, which can be grouped into three categories: geometric, intensity, and texture features. Second, all these features were optimized by principal component analysis algorithm to generate a compact and informative feature cluster. This cluster was used as input for developing and optimizing support vector machine (SVM) based classifiers, which indicated the likelihood of receiving suboptimal cytoreduction after the NACT treatment. Two different kernels for SVM algorithm were explored and compared. A total of 42 ovarian cancer cases were retrospectively collected to validate the scheme. A nested leave-one-out cross-validation framework was adopted for model performance assessment.

RESULTS

The results demonstrated that the model with a Gaussian radial basis function kernel SVM yielded an AUC (area under the ROC [receiver characteristic operation] curve) of 0.806 ± 0.078. Meanwhile, this model achieved overall accuracy (ACC) of 83.3%, positive predictive value (PPV) of 81.8%, and negative predictive value (NPV) of 83.9%.

CONCLUSION

This study provides meaningful information for the development of radiomics based image markers in NACT treatment outcome prediction.

Radiomics for differentiation of gliomas from primary central nervous system lymphomas: a systematic review and meta-analysis

Background and objective

Numerous radiomics-based models have been proposed to discriminate between central nervous system (CNS) gliomas and primary central nervous system lymphomas (PCNSLs). Given the heterogeneity of the existing models, we aimed to define their overall performance and identify the most critical variables to pilot future algorithms.

Methods

A systematic review of the literature and a meta-analysis were conducted, encompassing 12 studies and a total of 1779 patients, focusing on radiomics to differentiate gliomas from PCNSLs. A comprehensive literature search was performed through PubMed, Ovid MEDLINE, Ovid EMBASE, Web of Science, and Scopus databases. Overall sensitivity (SEN) and specificity (SPE) were estimated. Event rates were pooled using a random-effects meta-analysis, and the heterogeneity was assessed using the χ2 test.

Results

The overall SEN and SPE for differentiation between CNS gliomas and PCNSLs were 88% (95% CI = 0.83 - 0.91) and 87% (95% CI = 0.83 - 0.91), respectively. The best-performing features were the ones extracted from the Gray Level Run Length Matrix (GLRLM; ACC 97%), followed by those obtained from the Neighboring Gray Tone Difference Matrix (NGTDM; ACC 93%), and shape-based features (ACC 91%). The 18F-FDG-PET/CT was the best-performing imaging modality (ACC 97%), followed by the MRI CE-T1W (ACC 87% - 95%). Most studies applied a cross-validation analysis (92%).

Conclusion

The current SEN and SPE of radiomics to discriminate CNS gliomas from PCNSLs are high, making radiomics a helpful method to differentiate these tumor types. The best-performing features are the GLRLM, NGTDM, and shape-based features. The 18F-FDG-PET/CT imaging modality is the best-performing, while the MRI CE-T1W is the most used.

Penalized deep partially linear cox models with application to CT scans of lung cancer patients

Lung cancer is a leading cause of cancer mortality globally, highlighting the importance of understanding its mortality risks to design effective patient-centered therapies. The National Lung Screening Trial (NLST) employed computed tomography texture analysis, which provides objective measurements of texture patterns on CT scans, to quantify the mortality risks of lung cancer patients. Partially linear Cox models have gained popularity for survival analysis by dissecting the hazard function into parametric and nonparametric components, allowing for the effective incorporation of both well-established risk factors (such as age and clinical variables) and emerging risk factors (eg, image features) within a unified framework. However, when the dimension of parametric components exceeds the sample size, the task of model fitting becomes formidable, while nonparametric modeling grapples with the curse of dimensionality. We propose a novel Penalized Deep Partially Linear Cox Model (Penalized DPLC), which incorporates the smoothly clipped absolute deviation (SCAD) penalty to select important texture features and employs a deep neural network to estimate the nonparametric component of the model. We prove the convergence and asymptotic properties of the estimator and compare it to other methods through extensive simulation studies, evaluating its performance in risk prediction and feature selection. The proposed method is applied to the NLST study dataset to uncover the effects of key clinical and imaging risk factors on patients' survival. Our findings provide valuable insights into the relationship between these factors and survival outcomes.

Radiomic Biomarkers of Locoregional Recurrence: Prognostic Insights from Oral Cavity Squamous Cell Carcinoma preoperative CT scans

This study aimed to identify CT-based imaging biomarkers for locoregional recurrence (LR) in Oral Cavity Squamous Cell Carcinoma (OSCC) patients. Our study involved a retrospective review of 78 patients with OSCC who underwent surgical treatment at a single medical center. An approach involving feature selection and statistical model diagnostics was utilized to identify biomarkers. Two radiomics biomarkers, Large Dependence Emphasis (LDE) of the Gray Level Dependence Matrix (GLDM) and Long Run Emphasis (LRE) of the Gray Level Run Length Matrix (GLRLM) of the 3D Laplacian of Gaussian (LoG σ = 3), have demonstrated the capability to preoperatively distinguish patients with and without LR, exhibiting exceptional testing specificity (1.00) and sensitivity (0.82). The group with LRE > 2.99 showed a 3-year recurrence-free survival rate of 0.81, in contrast to 0.49 for the group with LRE ≤ 2.99. Similarly, the group with LDE > 120 showed a rate of 0.82, compared to 0.49 for the group with LDE ≤ 120. These biomarkers broaden our understanding of using radiomics to predict OSCC progression, enabling personalized treatment plans to enhance patient survival.

Radiomics in the characterization of lipid-poor adrenal adenomas at unenhanced CT: time to look beyond usual density metrics

OBJECTIVES

In this study, we developed a radiomic signature for the classification of benign lipid-poor adenomas, which may potentially help clinicians limit the number of unnecessary investigations in clinical practice. Indeterminate adrenal lesions of benign and malignant nature may exhibit different values of key radiomics features.

METHODS

Patients who had available histopathology reports and a non-contrast-enhanced CT scan were included in the study. Radiomics feature extraction was done after the adrenal lesions were contoured. The primary feature selection and prediction performance scores were calculated using the least absolute shrinkage and selection operator (LASSO). To eliminate redundancy, the best-performing features were further examined using the Pearson correlation coefficient, and new predictive models were created.

RESULTS

This investigation covered 50 lesions in 48 patients. After LASSO-based radiomics feature selection, the test dataset's 30 iterations of logistic regression models produced an average performance of 0.72. The model with the best performance, made up of 13 radiomics features, had an AUC of 0.99 in the training phase and 1.00 in the test phase. The number of features was lowered to 5 after performing Pearson's correlation to prevent overfitting. The final radiomic signature trained a number of machine learning classifiers, with an average AUC of 0.93.

CONCLUSIONS

Including more radiomics features in the identification of adenomas may improve the accuracy of NECT and reduce the need for additional imaging procedures and clinical workup, according to this and other recent radiomics studies that have clear points of contact with current clinical practice.

CLINICAL RELEVANCE STATEMENT

The study developed a radiomic signature using unenhanced CT scans for classifying lipid-poor adenomas, potentially reducing unnecessary investigations that scored a final accuracy of 93%.

KEY POINTS

• Radiomics has potential for differentiating lipid-poor adenomas and avoiding unnecessary further investigations. • Quadratic mean, strength, maximum 3D diameter, volume density, and area density are promising predictors for adenomas. • Radiomics models reach high performance with average AUC of 0.95 in the training phase and 0.72 in the test phase.

Quantitation of Oncologic Image Features for Radiomic Analyses in PET

Radiomics is an emerging and exciting field of study involving the extraction of many quantitative features from radiographic images. Positron emission tomography (PET) images are used in cancer diagnosis and staging. Utilizing radiomics on PET images can better quantify the spatial relationships between image voxels and generate more consistent and accurate results for diagnosis, prognosis, treatment, etc. This chapter gives the general steps a researcher would take to extract PET radiomic features from medical images and properly develop models to implement.

Principal Component Analysis Applied to Radiomics Data: Added Value for Separating Benign from Malignant Solitary Pulmonary Nodules

Here, we report on the added value of principal component analysis applied to a dataset of texture features derived from 39 solitary pulmonary lung nodule (SPN) lesions for the purpose of differentiating benign from malignant lesions, as compared to the use of SUVmax alone. Texture features were derived using the LIFEx software. The eight best-performing first-, second-, and higher-order features for separating benign from malignant nodules, in addition to SUVmax (MaximumGreyLevelSUVbwIBSI184IY), were included for PCA. Two principal components (PCs) were retained, of which the contributions to the total variance were, respectively, 87.6% and 10.8%. When included in a logistic binomial regression analysis, including age and gender as covariates, both PCs proved to be significant predictors for the underlying benign or malignant character of the lesions under study (p = 0.009 for the first PC and 0.020 for the second PC). As opposed to SUVmax alone, which allowed for the accurate classification of 69% of the lesions, the regression model including both PCs allowed for the accurate classification of 77% of the lesions. PCs derived from PCA applied on selected texture features may allow for more accurate characterization of SPN when compared to SUVmax alone.

A Critical Analysis of the Robustness of Radiomics to Variations in Segmentation Methods in 18F-PSMA-1007 PET Images of Patients Affected by Prostate Cancer

BACKGROUND

Radiomics shows promising results in supporting the clinical decision process, and much effort has been put into its standardization, thus leading to the Imaging Biomarker Standardization Initiative (IBSI), that established how radiomics features should be computed. However, radiomics still lacks standardization and many factors, such as segmentation methods, limit study reproducibility and robustness.

AIM

We investigated the impact that three different segmentation methods (manual, thresholding and region growing) have on radiomics features extracted from 18F-PSMA-1007 Positron Emission Tomography (PET) images of 78 patients (43 Low Risk, 35 High Risk). Segmentation was repeated for each patient, thus leading to three datasets of segmentations. Then, feature extraction was performed for each dataset, and 1781 features (107 original, 930 Laplacian of Gaussian (LoG) features, 744 wavelet features) were extracted. Feature robustness and reproducibility were assessed through the intra class correlation coefficient (ICC) to measure agreement between the three segmentation methods. To assess the impact that the three methods had on machine learning models, feature selection was performed through a hybrid descriptive-inferential method, and selected features were given as input to three classifiers, K-Nearest Neighbors (KNN), Support Vector Machines (SVM), Linear Discriminant Analysis (LDA), Random Forest (RF), AdaBoost and Neural Networks (NN), whose performance in discriminating between low-risk and high-risk patients have been validated through 30 times repeated five-fold cross validation.

CONCLUSIONS

Our study showed that segmentation methods influence radiomics features and that Shape features were the least reproducible (average ICC: 0.27), while GLCM features the most reproducible. Moreover, feature reproducibility changed depending on segmentation type, resulting in 51.18% of LoG features exhibiting excellent reproducibility (range average ICC: 0.68-0.87) and 47.85% of wavelet features exhibiting poor reproducibility that varied between wavelet sub-bands (range average ICC: 0.34-0.80) and resulted in the LLL band showing the highest average ICC (0.80). Finally, model performance showed that region growing led to the highest accuracy (74.49%), improved sensitivity (84.38%) and AUC (79.20%) in contrast with manual segmentation.

Identification of CT-based non-invasive radiomic biomarkers for overall survival prediction in oral cavity squamous cell carcinoma

This study addresses the limited non-invasive tools for Oral Cavity Squamous Cell Carcinoma (OSCC) survival prediction by identifying Computed Tomography (CT)-based biomarkers to improve prognosis prediction. A retrospective analysis was conducted on data from 149 OSCC patients, including CT radiomics and clinical information. An ensemble approach involving correlation analysis, score screening, and the Sparse-L1 algorithm was used to select functional features, which were then used to build Cox Proportional Hazards models (CPH). Our CPH achieved a 0.70 concordance index in testing. The model identified two CT-based radiomics features, Gradient-Neighboring-Gray-Tone-Difference-Matrix-Strength (GNS) and normalized-Wavelet-LLL-Gray-Level-Dependence-Matrix-Large-Dependence-High-Gray-Level-Emphasis (HLE), as well as stage and alcohol usage, as survival biomarkers. The GNS group with values above 14 showed a hazard ratio of 0.12 and a 3-year survival rate of about 90%. Conversely, the GNS group with values less than or equal to 14 had a 49% survival rate. For normalized HLE, the high-end group (HLE > - 0.415) had a hazard ratio of 2.41, resulting in a 3-year survival rate of 70%, while the low-end group (HLE ≤ - 0.415) had a 36% survival rate. These findings contribute to our knowledge of how radiomics can be used to predict the outcome so that treatment plans can be tailored for patients people with OSCC to improve their survival.

Imaging phenotypes predict overall survival in glioma more accurate than basic demographic and cell mutation profiles

BACKGROUND

Magnetic resonance imaging (MRI), digital pathology imaging (PATH), demographics, and IDH mutation status predict overall survival (OS) in glioma. Identifying and characterizing predictive features in the different modalities may improve OS prediction accuracy.

PURPOSE

To evaluate the OS prediction accuracy of combinations of prognostic markers in glioma patients.

MATERIALS AND METHODS

Multi-contrast MRI, comprising T1-weighted, T1-weighted post-contrast, T2-weighted, T2 fluid-attenuated-inversion-recovery, and pathology images from glioma patients (n = 160) were retrospectively collected (1983-2008) from TCGA alongside age and sex. Phenotypic profiling of tumors was performed by quantifying the radiographic and histopathologic descriptors extracted from the delineated region-of-interest in MRI and PATH images. A Cox proportional hazard model was trained with the MRI and PATH features, IDH mutation status, and basic demographic variables (age and sex) to predict OS. The performance was evaluated in a split-train-test configuration using the concordance-index, computed between the predicted risk score and observed OS.

RESULTS

The average age of patients was 51.2years (women: n = 77, age-range=18-84years; men: n = 83, age-range=21-80years). The median OS of the participants was 494.5 (range,3-4752), 481 (range,7-4752), and 524.5 days (range,3-2869), respectively, in complete dataset, training, and test datasets. The addition of MRI or PATH features improved prediction of OS when compared to models based on age, sex, and mutation status alone or their combination (p < 0.001). The full multi-omics model integrated MRI, PATH, clinical, and genetic profiles and predicted the OS best (c-index= 0.87).

CONCLUSION

The combination of imaging, genetic, and clinical profiles leads to a more accurate prognosis than the clinical and/or mutation status.

CT-based Radiogenomics Framework for COVID-19 Using ACE2 Imaging Representations

Coronavirus disease 2019 (COVID-19) is caused by Severe Acute Respiratory Syndrome Coronavirus 2 which enters the body via the angiotensin-converting enzyme 2 (ACE2) and altering its gene expression. Altered ACE2 plays a crucial role in the pathogenesis of COVID-19. Gene expression profiling, however, is invasive and costly, and is not routinely performed. In contrast, medical imaging such as computed tomography (CT) captures imaging features that depict abnormalities, and it is widely available. Computerized quantification of image features has enabled 'radiogenomics', a research discipline that identifies image features that are associated with molecular characteristics. Radiogenomics between ACE2 and COVID-19 has yet to be done primarily due to the lack of ACE2 expression data among COVID-19 patients. Similar to COVID-19, patients with lung adenocarcinoma (LUAD) exhibit altered ACE2 expression and, LUAD data are abundant. We present a radiogenomics framework to derive image features (ACE2-RGF) associated with ACE2 expression data from LUAD. The ACE2-RGF was then used as a surrogate biomarker for ACE2 expression. We adopted conventional feature selection techniques including ElasticNet and LASSO. Our results show that: i) the ACE2-RGF encoded a distinct collection of image features when compared to conventional techniques, ii) the ACE2-RGF can classify COVID-19 from normal subjects with a comparable performance to conventional feature selection techniques with an AUC of 0.92, iii) ACE2-RGF can effectively identify patients with critical illness with an AUC of 0.85. These findings provide unique insights for automated COVID-19 analysis and future research.

Preoperative 18F-FDG PET/CT and CT radiomics for identifying aggressive histopathological subtypes in early stage lung adenocarcinoma

Lung adenocarcinoma (ADC) is the most common non-small cell lung cancer. Surgical resection is the primary treatment for early-stage lung ADC while lung-sparing surgery is an alternative for non-aggressive cases. Identifying histopathologic subtypes before surgery helps determine the optimal surgical approach. Predominantly solid or micropapillary (MIP) subtypes are aggressive and associated with a higher likelihood of recurrence and metastasis and lower survival rates. This study aims to non-invasively identify these aggressive subtypes using preoperative 18F-FDG PET/CT and diagnostic CT radiomics analysis. We retrospectively studied 119 patients with stage I lung ADC and tumors ≤ 2 cm, where 23 had aggressive subtypes (18 solid and 5 MIPs). Out of 214 radiomic features from the PET/CT and CT scans and 14 clinical parameters, 78 significant features (3 CT and 75 PET features) were identified through univariate analysis and hierarchical clustering with minimized feature collinearity. A combination of Support Vector Machine classifier and Least Absolute Shrinkage and Selection Operator built predictive models. Ten iterations of 10-fold cross-validation (10 ×10-fold CV) evaluated the model. A pair of texture feature (PET GLCM Correlation) and shape feature (CT Sphericity) emerged as the best predictor. The radiomics model significantly outperformed the conventional predictor SUVmax (accuracy: 83.5% vs. 74.7%, p = 9e-9) and identified aggressive subtypes by evaluating FDG uptake in the tumor and tumor shape. It also demonstrated a high negative predictive value of 95.6% compared to SUVmax (88.2%, p = 2e-10). The proposed radiomics approach could reduce unnecessary extensive surgeries for non-aggressive subtype patients, improving surgical decision-making for early-stage lung ADC patients.

CT Imaging With Machine Learning for Predicting Progression to COPD in Individuals at Risk

BACKGROUND

Identifying individuals at risk of progressing to COPD may allow for initiation of treatment to potentially slow the progression of the disease or the selection of subgroups for discovery of novel interventions.

RESEARCH QUESTION

Does the addition of CT imaging features, texture-based radiomic features, and established quantitative CT scan to conventional risk factors improve the performance for predicting progression to COPD in individuals who smoke with machine learning?

STUDY DESIGN AND METHODS

Participants at risk (individuals who currently or formerly smoked, without COPD) from the Canadian Cohort Obstructive Lung Disease (CanCOLD) population-based study underwent CT imaging at baseline and spirometry at baseline and follow-up. Various combinations of CT scan features, texture-based CT scan radiomics (n = 95), and established quantitative CT scan (n = 8), as well as demographic (n = 5) and spirometry (n = 3) measurements, with machine learning algorithms were evaluated to predict progression to COPD. Performance metrics included the area under the receiver operating characteristic curve (AUC) to evaluate the models. DeLong test was used to compare the performance of the models.

RESULTS

Among the 294 at-risk participants who were evaluated (mean age, 65.6 ± 9.2 years; 42% female; mean pack-years, 17.9 ± 18.7), 52 participants (23.7%) in the training data set and 17 participants (23.0%) in the testing data set progressed to spirometric COPD at follow-up (2.5 ± 0.9 years from baseline). Compared with machine learning models with demographics alone (AUC, 0.649), the addition of CT imaging features to demographics (AUC, 0.730; P < .05) or CT imaging features and spirometry to demographics (AUC, 0.877; P < .05) significantly improved the performance for predicting progression to COPD.

INTERPRETATION

Heterogeneous structural changes occur in the lungs of individuals at risk that can be quantified using CT imaging features, and evaluation of these features together with conventional risk factors improves performance for predicting progression to COPD.

Performance of hippocampal radiomics models based on T2-FLAIR images in mesial temporal lobe epilepsy with hippocampal sclerosis

PURPOSE

Preoperative identification of hippocampal sclerosis (HS) is crucial to successful surgery for mesial temporal lobe epilepsy (MTLE). We aimed to investigate the diagnostic performance of hippocampal radiomics models based on T2 fluid-attenuated inversion recovery (FLAIR) images in MTLE with HS.

METHODS

We analysed 210 cases, including 172 HS pathology-confirmed cases (100 magnetic resonance imaging [MRI]-positive cases [MRI + HS], 72 MRI-negative HS cases [MRI - HS]), and 38 healthy controls (HC). The hippocampus was delineated slice by slice on an oblique coronal plane by a T2-FLAIR sequence, perpendicular to the hippocampus's long axis, to obtain a three-dimensional region of interest. Radiomics were processed using Artificial Intelligence Kit software; logistic regression radiomics models were constructed. The model evaluation indexes included the area under the curve (AUC), accuracy, sensitivity, and specificity.

RESULTS

The respective AUC, accuracy, sensitivity, and specificity were 0.863, 81.4%, 78.0%, and 84.6% between the MRI - HS and HC groups in the training set and 0.855, 75.0%, 68.2%, and 81.8% in the test set; 0.975, 95.0%, 92.9%, and 98.0% between the MRI + HS and HC groups in the training set and 0.954, 88.7%, 90.0%, and 87.0% in the test set; and 0.912, 84.3%, 83.3%, and 86.5% between the MTLE and HC groups in the training set and 0.854, 79.7%, 80.8%, and 77.3% in the test set. The AUC values of the comparative radiomics models were > 0.85, indicating good diagnostic efficiency.

CONCLUSION

The hippocampal radiomics models based on T2-FLAIR images can help diagnose MTLE with HS. They can be used as biological markers for MTLE diagnosis.

Topology-based radiomic features for prediction of parotid gland cancer malignancy grade in magnetic resonance images

PURPOSE

The malignancy grades of parotid gland cancer (PGC) have been assessed for a decision of treatment policies. Therefore, we have investigated the feasibility of topology-based radiomic features for the prediction of parotid gland cancer (PGC) malignancy grade in magnetic resonance (MR) images.

MATERIALS AND METHODS

Two-dimensional T1- and T2-weighted MR images of 39 patients with PGC were selected for this study. Imaging properties of PGC can be quantified using the topology, which could be useful for assessing the number of the k-dimensional holes or heterogeneity in PGC regions using invariants of the Betti numbers. Radiomic signatures were constructed from 41,472 features obtained after a harmonization using an elastic net model. PGC patients were stratified using a logistic classification into low/intermediate- and high-grade malignancy groups. The training data were increased by four times to avoid the overfitting problem using a synthetic minority oversampling technique. The proposed approach was assessed using a 4-fold cross-validation test.

RESULTS

The highest accuracy of the proposed approach was 0.975 for the validation cases, whereas that of the conventional approach was 0.694.

CONCLUSION

This study indicated that topology-based radiomic features could be feasible for the noninvasive prediction of the malignancy grade of PGCs.