Personalized Risk Analysis to Improve the Psychological Resilience of Women Undergoing Treatment for Breast Cancer: Development of a Machine Learning-Driven Clinical Decision Support Tool.

BACKGROUND: Health professionals are often faced with the need to identify women at risk of manifesting poor psychological resilience following the diagnosis and treatment of breast cancer. Machine learning algorithms are increasingly used to support clinical decision support (CDS) tools in helping health professionals identify women who are at risk of adverse well-being outcomes and plan customized psychological interventions for women at risk. Clinical flexibility, cross-validated performance accuracy, and model explainability permitting person-specific identification of risk factors are highly desirable features of such tools. OBJECTIVE: This study aimed to develop and cross-validate machine learning models designed to identify breast cancer survivors at risk of poor overall mental health and global quality of life and identify potential targets of personalized psychological interventions according to an extensive set of clinical recommendations. METHODS: A set of 12 alternative models was developed to improve the clinical flexibility of the CDS tool. All models were validated using longitudinal data from a prospective, multicenter clinical pilot at 5 major oncology centers in 4 countries (Italy, Finland, Israel, and Portugal; the Predicting Effective Adaptation to Breast Cancer to Help Women to BOUNCE Back [BOUNCE] project). A total of 706 patients with highly treatable breast cancer were enrolled shortly after diagnosis and before the onset of oncological treatments and were followed up for 18 months. An extensive set of demographic, lifestyle, clinical, psychological, and biological variables measured within 3 months after enrollment served as predictors. Rigorous feature selection isolated key psychological resilience outcomes that could be incorporated into future clinical practice. RESULTS: Balanced random forest classifiers were successful at predicting well-being outcomes, with accuracies ranging between 78% and 82% (for 12-month end points after diagnosis) and between 74% and 83% (for 18-month end points after diagnosis). Explainability and interpretability analyses built on the best-performing models were used to identify potentially modifiable psychological and lifestyle characteristics that, if addressed systematically in the context of personalized psychological interventions, would be most likely to promote resilience for a given patient. CONCLUSIONS: Our results highlight the clinical utility of the BOUNCE modeling approach by focusing on resilience predictors that can be readily available to practicing clinicians at major oncology centers. The BOUNCE CDS tool paves the way for personalized risk assessment methods to identify patients at high risk of adverse well-being outcomes and direct valuable resources toward those most in need of specialized psychological interventions.

Magnetization transfer imaging to assess tumour response after chemoradiotherapy in rectal cancer.

PURPOSE: Single-slice magnetization transfer (MT) imaging has shown promising results for evaluating post-radiation fibrosis. The study aim was to evaluate the value of multislice MT imaging to assess tumour response after chemoradiotherapy by comparing magnetization transfer ratios (MTR) with histopathological tumour regression grade (TRG). MATERIALS AND METHODS: Thirty patients with locally advanced rectal cancer (cT3-4 and/or cN2) underwent routine restaging MRI 8 weeks post-chemoradiotherapy, including multislice MT-sequence, covering the entire tumour bed. Two independent readers delineated regions of interest on MTR maps, covering all potential remaining tumour and fibrotic areas. Mean MTR and histogram parameters (minimum, maximum, median, standard deviation, skewness, kurtosis, and 5-30-70-95th percentiles) were calculated. Reference standard was histological TRG1-2 (good response) and TRG3-5 (poor response). RESULTS: 24/30 patients were male; mean age was 67.7 ± 10.8 years. Mean MTR rendered AUCs of 0.65 (reader1) and 0.87 (reader2) to differentiate between TRG1-2 versus TRG3-5. Best results were obtained for 95(th) percentile (AUC 0.75- 0.88). Interobserver agreement was moderate (ICC 0.50) for mean MTR and good (ICC 0.80) for 95(th) percentile. CONCLUSIONS: MT imaging is a promising tool to assess tumour response post-chemoradiotherapy in rectal cancer. Particularly, 95(th) percentile results in AUCs up to 0.88 to discriminate a good tumour response. KEY POINTS: • The mean MTR can differentiate between good and poor responders after chemoradiation. • In addition to measurement of the mean value, histogram analyses can be beneficial. • The histogram parameter 95 (th) percentile can reach AUCs of 0.75-0.88.

Unveiling the Power of Model-Agnostic Multiscale Analysis for Enhancing Artificial Intelligence Models in Breast Cancer Histopathology Images.

Developing AI models for digital pathology has traditionally relied on single-scale analysis of histopathology slides. However, a whole slide image is a rich digital representation of the tissue, captured at various magnification levels. Limiting our analysis to a single scale overlooks critical information, spanning from intricate high-resolution cellular details to broad low-resolution tissue structures. In this study, we propose a model-agnostic multiresolution feature aggregation framework tailored for the analysis of histopathology slides in the context of breast cancer, on a multicohort dataset of 2038 patient samples. We have adapted 9 state-of-the-art multiple instance learning models on our multi-scale methodology and evaluated their performance on grade prediction, TP53 mutation status prediction and survival prediction. The results prove the dominance of the multiresolution methodology, and specifically, concatenating or linearly transforming via a learnable layer the feature vectors of image patches from a high (20x) and low (10x) magnification factors achieve improved performance for all prediction tasks across domain-specific and imagenet-based features. On the contrary, the performance of uniresolution baseline models was not consistent across domain-specific and imagenet-based features. Moreover, we shed light on the inherent inconsistencies observed in models trained on whole-tissue-sections when validated against biopsy-based datasets. Despite these challenges, our findings underscore the superiority of multiresolution analysis over uniresolution methods. Finally, cross-scale analysis also benefits the explainability aspects of attention-based architectures, since one can extract attention maps at the tissue- and cell-levels, improving the interpretation of the model's decision.

LKB1 Loss Correlates with STING Loss and, in Cooperation with ß-Catenin Membranous Loss, Indicates Poor Prognosis in Patients with Operable Non-Small Cell Lung Cancer.

To investigate the incidence and prognostically significant correlations and cooperations of LKB1 loss of expression in non-small cell lung cancer (NSCLC), surgical specimens from 188 metastatic and 60 non-metastatic operable stage I-IIIA NSCLC patients were analyzed to evaluate their expression of LKB1 and pAMPK proteins in relation to various processes. The investigated factors included antitumor immunity response regulators STING and PD-L1; pro-angiogenic, EMT and cell cycle targets, as well as metastasis-related (VEGFC, PDGFRα, PDGFRß, p53, p16, Cyclin D1, ZEB1, CD24) targets; and cell adhesion (ß-catenin) molecules. The protein expression levels were evaluated via immunohistochemistry; the RNA levels of LKB1 and NEDD9 were evaluated via PCR, while KRAS exon 2 and BRAFV600E mutations were evaluated by Sanger sequencing. Overall, loss of LKB1 protein expression was observed in 21% (51/248) patients and correlated significantly with histotype (p < 0.001), KRAS mutations (p < 0.001), KC status (concomitant KRAS mutation and p16 downregulation) (p < 0.001), STING loss (p < 0.001), and high CD24 expression (p < 0.001). STING loss also correlated significantly with loss of LKB1 expression in the metastatic setting both overall (p = 0.014) and in lung adenocarcinomas (LUACs) (p = 0.005). Additionally, LKB1 loss correlated significantly with a lack of or low ß-catenin membranous expression exclusively in LUACs, both independently of the metastatic status (p = 0.019) and in the metastatic setting (p = 0.007). Patients with tumors yielding LKB1 loss and concomitant nonexistent or low ß-catenin membrane expression experienced significantly inferior median overall survival of 20.50 vs. 52.99 months; p < 0.001 as well as significantly greater risk of death (HR: 3.32, 95% c.i.: 1.71-6.43; p <0.001). Our findings underscore the impact of the synergy of LKB1 with STING and ß-catenin in NSCLC, in prognosis.

A segmentation-based method improving the performance of N4 bias field correction on T2weighted MR imaging data of the prostate.

Magnetic Resonance (MR) images suffer from spatial inhomogeneity, known as bias field corruption. The N4ITK filter is a state-of-the-art method used for correcting the bias field to optimize MR-based quantification. In this study, a novel approach is presented to quantitatively evaluate the performance of N4 bias field correction for pelvic prostate imaging. An exploratory analysis, regarding the different values of convergence threshold, shrink factor, fitting level, number of iterations and use of mask, is performed to quantify the performance of N4 filter in pelvic MR images. The performance of a total of 240 different N4 configurations is examined using the Full Width at Half Maximum (FWHM) of the segmented periprostatic fat distribution as evaluation metric. Phantom T2weighted images were used to assess the performance of N4 for a uniform test tissue mimicking material, excluding factors such as patient related susceptibility and anatomy heterogeneity. Moreover, 89 and 204 T2weighted patient images from two public datasets acquired by scanners with a combined surface and endorectal coil at 1.5 T and a surface coil at 3 T, respectively, were utilized and corrected with a variable set of N4 parameters. Furthermore, two external public datasets were used to validate the performance of the N4 filter in T2weighted patient images acquired by various scanning conditions with different magnetic field strengths and coils. The results show that the set of N4 parameters, converging to optimal representations of fat in the image, were: convergence threshold 0.001, shrink factor 2, fitting level 6, number of iterations 100 and the use of default mask for prostate images acquired by a combined surface and endorectal coil at both 1.5 T and 3 T. The corresponding optimal N4 configuration for MR prostate images acquired by a surface coil at 1.5 T or 3 T was: convergence threshold 0.001, shrink factor 2, fitting level 5, number of iterations 25 and the use of default mask. Hence, periprostatic fat segmentation can be used to define the optimal settings for achieving T2weighted prostate images free from bias field corruption to provide robust input for further analysis.

Multiresolution Self-Supervised Feature Integration via Attention Multiple Instance Learning for Histopathology Analysis.

Digital histopathology image analysis of tumor tissue sections has seen great research interest for automating standard diagnostic tasks, but also for developing novel prognostic biomarkers. However, research has mainly been focused on developing uniresolution models, capturing either high-resolution cellular features or low-resolution tissue architectural features. In addition, in the patch-based weakly-supervised training of deep learning models, the features which represent the intratumoral heterogeneity are lost. In this study, we propose a multiresolution attention-based multiple instance learning framework that can capture cellular and contextual features from the whole tissue for predicting patient-level outcomes. Several basic mathematical operations were examined for integrating multiresolution features, i.e. addition, mean, multiplication and concatenation. The proposed multiplication-based multiresolution model performed the best (AUC=0.864), while all multiresolution models outperformed the uniresolution baseline models (AUC=0.669, 0.713) for breast-cancer grading. (Implementation: https://github.com/tsikup/multiresolution-clam).

Personalized prediction of one-year mental health deterioration using adaptive learning algorithms: a multicenter breast cancer prospective study.

Identifying individual patient characteristics that contribute to long-term mental health deterioration following diagnosis of breast cancer (BC) is critical in clinical practice. The present study employed a supervised machine learning pipeline to address this issue in a subset of data from a prospective, multinational cohort of women diagnosed with stage I-III BC with a curative treatment intention. Patients were classified as displaying stable HADS scores (Stable Group; n = 328) or reporting a significant increase in symptomatology between BC diagnosis and 12 months later (Deteriorated Group; n = 50). Sociodemographic, life-style, psychosocial, and medical variables collected on the first visit to their oncologist and three months later served as potential predictors of patient risk stratification. The flexible and comprehensive machine learning (ML) pipeline used entailed feature selection, model training, validation and testing. Model-agnostic analyses aided interpretation of model results at the variable- and patient-level. The two groups were discriminated with a high degree of accuracy (Area Under the Curve = 0.864) and a fair balance of sensitivity (0.85) and specificity (0.87). Both psychological (negative affect, certain coping with cancer reactions, lack of sense of control/positive expectations, and difficulties in regulating negative emotions) and biological variables (baseline percentage of neutrophils, thrombocyte count) emerged as important predictors of mental health deterioration in the long run. Personalized break-down profiles revealed the relative impact of specific variables toward successful model predictions for each patient. Identifying key risk factors for mental health deterioration is an essential first step toward prevention. Supervised ML models may guide clinical recommendations toward successful illness adaptation.

New Insights in the Era of Clinical Biomarkers as Potential Predictors of Systemic Therapy-Induced Cardiotoxicity in Women with Breast Cancer: A Systematic Review.

Cardiotoxicity induced by breast cancer therapies is a potentially serious complication associated with the use of various breast cancer therapies. Prediction and better management of cardiotoxicity in patients receiving chemotherapy is of critical importance. However, the management of cancer therapy-related cardiac dysfunction (CTRCD) lacks clinical evidence and is based on limited clinical studies. AIM: To provide an overview of existing and potentially novel biomarkers that possess a promising predictive value for the early and late onset of CTRCD in the clinical setting. METHODS: A systematic review of published studies searching for promising biomarkers for the prediction of CTRCD in patients with breast cancer was undertaken according to PRISMA guidelines. A search strategy was performed using PubMed, Google Scholar, and Scopus for the period 2013-2023. All subjects were >18 years old, diagnosed with breast cancer, and received breast cancer therapies. RESULTS: The most promising biomarkers that can be used for the development of an alternative risk cardiac stratification plan for the prediction and/or early detection of CTRCD in patients with breast cancer were identified. CONCLUSIONS: We highlighted the new insights associated with the use of currently available biomarkers as a standard of care for the management of CTRCD and identified potentially novel clinical biomarkers that could be further investigated as promising predictors of CTRCD.

Simulating radiotherapy effect in high-grade glioma by using diffusive modeling and brain atlases.

Applying diffusive models for simulating the spatiotemporal change of concentration of tumour cells is a modern application of predictive oncology. Diffusive models are used for modelling glioblastoma, the most aggressive type of glioma. This paper presents the results of applying a linear quadratic model for simulating the effects of radiotherapy on an advanced diffusive glioma model. This diffusive model takes into consideration the heterogeneous velocity of glioma in gray and white matter and the anisotropic migration of tumor cells, which is facilitated along white fibers. This work uses normal brain atlases for extracting the proportions of white and gray matter and the diffusion tensors used for anisotropy. The paper also presents the results of applying this glioma model on real clinical datasets.

Harmonization Strategies in Multicenter MRI-Based Radiomics.

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

Dissecting Tumor-Immune Microenvironment in Breast Cancer at a Spatial and Multiplex Resolution.

The tumor immune microenvironment (TIME) is an important player in breast cancer pathophysiology. Surrogates for antitumor immune response have been explored as predictive biomarkers to immunotherapy, though with several limitations. Immunohistochemistry for programmed death ligand 1 suffers from analytical problems, immune signatures are devoid of spatial information and histopathological evaluation of tumor infiltrating lymphocytes exhibits interobserver variability. Towards improved understanding of the complex interactions in TIME, several emerging multiplex in situ methods are being developed and gaining much attention for protein detection. They enable the simultaneous evaluation of multiple targets in situ, detection of cell densities/subpopulations as well as estimations of functional states of immune infiltrate. Furthermore, they can characterize spatial organization of TIME-by cell-to-cell interaction analyses and the evaluation of distribution within different regions of interest and tissue compartments-while digital imaging and image analysis software allow for reproducibility of the various assays. In this review, we aim to provide an overview of the different multiplex in situ methods used in cancer research with special focus on breast cancer TIME at the neoadjuvant, adjuvant and metastatic setting. Spatial heterogeneity of TIME and importance of longitudinal evaluation of TIME changes under the pressure of therapy and metastatic progression are also addressed.

Pollen Grain Classification Based on Ensemble Transfer Learning on the Cretan Pollen Dataset.

Pollen identification is an important task for the botanical certification of honey. It is performed via thorough microscopic examination of the pollen present in honey; a process called melissopalynology. However, manual examination of the images is hard, time-consuming and subject to inter- and intra-observer variability. In this study, we investigated the applicability of deep learning models for the classification of pollen-grain images into 20 pollen types, based on the Cretan Pollen Dataset. In particular, we applied transfer and ensemble learning methods to achieve an accuracy of 97.5%, a sensitivity of 96.9%, a precision of 97%, an F1 score of 96.89% and an AUC of 0.9995. However, in a preliminary case study, when we applied the best-performing model on honey-based pollen-grain images, we found that it performed poorly; only 0.02 better than random guessing (i.e., an AUC of 0.52). This indicates that the model should be further fine-tuned on honey-based pollen-grain images to increase its effectiveness on such data.

Diffusion Weighted Imaging in the Assessment of Tumor Grade in Endometrial Cancer Based on Intravoxel Incoherent Motion MRI.

The aim of this study is to investigate the possibility of predicting histological grade in patients with endometrial cancer on the basis of intravoxel incoherent motion (IVIM)-related histogram analysis parameters. This prospective study included 52 women with endometrial cancer (EC) who underwent MR imaging as initial staging in our hospital, allocated into low-grade (G1 and G2) and high-grade (G3) tumors according to the pathology reports. Regions of interest (ROIs) were drawn on the diffusion weighted images and apparent diffusion coefficient (ADC), true diffusivity (D), and perfusion fraction (f) using diffusion models were computed. Mean, median, skewness, kurtosis, and interquartile range (IQR) were calculated from the whole-tumor histogram. The IQR of the diffusion coefficient (D) was significantly lower in the low-grade tumors from that of the high-grade group with an adjusted p-value of less than 5% (0.048). The ROC curve analysis results of the statistically significant IQR of the D yielded an accuracy, sensitivity, and specificity of 74.5%, 70.1%, and 76.5% respectively, for discriminating low from high-grade tumors, with an optimal cutoff of 0.206 (×10-3 mm2/s) and an AUC of 75.4% (95% CI: 62.1 to 88.8). The IVIM modeling coupled with histogram analysis techniques is promising for preoperative differentiation between low- and high-grade EC tumors.

Trajectories and Predictors of Depression After Breast Cancer Diagnosis: A 1-year longitudinal study.

Being diagnosed with breast cancer (BC) can be a traumatic experience for patients who may experience symptoms of depression. In order to facilitate the prevention of such symptoms, it is crucial to understand how and why depressive symptoms emerge and evolve for each individual, from diagnosis through treatment and recovery. In the present work, data from a multicentric study of 706 BC patients followed for 12 months are analyzed. First, a trajectory-based unsupervised clustering based on K-means is performed to capture the dynamic patterns of change in patients' depressive symptoms after BC diagnosis and to identify distinct trajectory clusters. Then a supervised learning approach was employed to build a classification model of depression progression and to identify potential predictors. Patients were clustered into 4 groups: stable low, stable high, improving, and worsening depressive symptoms. In a nested cross-validation pipeline, the performance of the Support Vector Machine model for discriminating between "good" and "poor" progression was 0.78±0.05 in terms of AUC. Several psychological variables emerged as highly predictive of the evolution of depressive symptoms with the most important ones being negative affectivity and anxious preoccupation. Clinical Relevance-The findings of the present study may help clinicians tailor individualized psychological interventions aiming at alleviating the burden of these symptoms in women with breast cancer and improving their overall well-being.

Differentiating low from high-grade soft tissue sarcomas using post-processed imaging parameters derived from multiple DWI models.

PURPOSE: To investigate and histopathologically validate the role of model selection in the design of novel parametric meta-maps towards the discrimination of low from high-grade soft tissue sarcomas (STSs) using multiple Diffusion Weighted Imaging (DWI) models. METHODS: DWI data of 28 patients were quantified using the mono-exponential, bi-exponential, stretched-exponential and the diffusion kurtosis model. Akaike Weights (AW) were calculated from the corrected Akaike Information Criteria (AICc) to select the most suitable model for every pixel within the tumor volume. Pseudo-colorized classification maps were then generated to depict model suitability, hypothesizing that every single model underpins different tissue properties and cannot solely characterize the whole tumor. Single model parametric maps were turned into meta-maps using the classification map and a histological validation of the model suitability results was conducted on several subregions of different tumors. Several histogram metrics were calculated from all derived maps before and after model selection, statistical analysis was conducted using the Mann-Whitney U test, p-values were adjusted for multiple comparisons and performance of all statistically significant metrics was evaluated using the Receiver Operator Characteristic (ROC) analysis. RESULTS: The histologic analysis on several tumor subregions confirmed model suitability results on these areas. Only 3 histogram metrics, all derived from the meta-maps, were found to be statistically significant in differentiating low from high-grade STSs with an AUC higher than 89 %. CONCLUSION: Embedding model selection in the design of the diffusion parametric maps yields to histogram metrics of high discriminatory power in grading STSs.

Deep learning for diabetic retinopathy detection and classification based on fundus images: A review.

Diabetic Retinopathy is a retina disease caused by diabetes mellitus and it is the leading cause of blindness globally. Early detection and treatment are necessary in order to delay or avoid vision deterioration and vision loss. To that end, many artificial-intelligence-powered methods have been proposed by the research community for the detection and classification of diabetic retinopathy on fundus retina images. This review article provides a thorough analysis of the use of deep learning methods at the various steps of the diabetic retinopathy detection pipeline based on fundus images. We discuss several aspects of that pipeline, ranging from the datasets that are widely used by the research community, the preprocessing techniques employed and how these accelerate and improve the models' performance, to the development of such deep learning models for the diagnosis and grading of the disease as well as the localization of the disease's lesions. We also discuss certain models that have been applied in real clinical settings. Finally, we conclude with some important insights and provide future research directions.

Prediction of Poor Mental Health Following Breast Cancer Diagnosis Using Random Forests1.

Breast cancer diagnosis has been associated with poor mental health, with significant impairment of quality of life. In order to ensure support for successful adaptation to this illness, it is of paramount importance to identify the most prominent factors affecting well-being that allow for accurate prediction of mental health status across time. Here we exploit a rich set of clinical, psychological, socio-demographic and lifestyle data from a large multicentre study of patients recently diagnosed with breast cancer, in order to classify patients based on their mental health status and further identify potential predictors of such status. For this purpose, a supervised learning pipeline using cross-sectional data was implemented for the formulation of a classification scheme of mental health status 6 months after diagnosis. Model performance in terms of AUC ranged from 0.81± 0.04 to 0.90± 0.03. Several psychological variables, including initial levels of anxiety and depression, emerged as highly predictive of short-term mental health status of women diagnosed with breast cancer.

Multicenter DSC-MRI-Based Radiomics Predict IDH Mutation in Gliomas.

To address the current lack of dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI)-based radiomics to predict isocitrate dehydrogenase (IDH) mutations in gliomas, we present a multicenter study that featured an independent exploratory set for radiomics model development and external validation using two independent cohorts. The maximum performance of the IDH mutation status prediction on the validation set had an accuracy of 0.544 (Cohen's kappa: 0.145, F1-score: 0.415, area under the curve-AUC: 0.639, sensitivity: 0.733, specificity: 0.491), which significantly improved to an accuracy of 0.706 (Cohen's kappa: 0.282, F1-score: 0.474, AUC: 0.667, sensitivity: 0.6, specificity: 0.736) when dynamic-based standardization of the images was performed prior to the radiomics. Model explainability using local interpretable model-agnostic explanations (LIME) and Shapley additive explanations (SHAP) revealed potential intuitive correlations between the IDH-wildtype increased heterogeneity and the texture complexity. These results strengthened our hypothesis that DSC-MRI radiogenomics in gliomas hold the potential to provide increased predictive performance from models that generalize well and provide understandable patterns between IDH mutation status and the extracted features toward enabling the clinical translation of radiogenomics in neuro-oncology.

Radiomics and Machine Learning Can Differentiate Transient Osteoporosis from Avascular Necrosis of the Hip.

Differentiation between transient osteoporosis (TOH) and avascular necrosis (AVN) of the hip is a longstanding challenge in musculoskeletal radiology. The purpose of this study was to utilize MRI-based radiomics and machine learning (ML) for accurate differentiation between the two entities. A total of 109 hips with TOH and 104 hips with AVN were retrospectively included. Femoral heads and necks with segmented radiomics features were extracted. Three ML classifiers (XGboost, CatBoost and SVM) using 38 relevant radiomics features were trained on 70% and validated on 30% of the dataset. ML performance was compared to two musculoskeletal radiologists, a general radiologist and two radiology residents. XGboost achieved the best performance with an area under the curve (AUC) of 93.7% (95% CI from 87.7 to 99.8%) among ML models. MSK radiologists achieved an AUC of 90.6% (95% CI from 86.7% to 94.5%) and 88.3% (95% CI from 84% to 92.7%), respectively, similar to residents. The general radiologist achieved an AUC of 84.5% (95% CI from 80% to 89%), significantly lower than of XGboost (p = 0.017). In conclusion, radiomics-based ML achieved a performance similar to MSK radiologists and significantly higher compared to general radiologists in differentiating between TOH and AVN.

A machine learning-based pipeline for modeling medical, socio-demographic, lifestyle and self-reported psychological traits as predictors of mental health outcomes after breast cancer diagnosis: An initial effort to define resilience effects.

Displaying resilience following a diagnosis of breast cancer is crucial for successful adaptation to illness, well-being, and health outcomes. Several theoretical and computational models have been proposed toward understanding the complex process of illness adaptation, involving a large variety of patient sociodemographic, lifestyle, medical, and psychological characteristics. To date, conventional multivariate statistical methods have been used extensively to model resilience. In the present work we describe a computational pipeline designed to identify the most prominent predictors of mental health outcomes following breast cancer diagnosis. A machine learning framework was developed and tested on the baseline data (recorded immediately post diagnosis) from an ongoing prospective, multinational study. This fully annotated dataset includes socio-demographic, lifestyle, medical and self-reported psychological characteristics of women recently diagnosed with breast cancer (N = 609). Nine different feature selection and cross-validated classification schemes were compared on their performance in classifying patients into low vs high depression symptom severity. Best-performing approaches involved a meta-estimator combined with a Support Vector Machines (SVMs) classification algorithm, exhibiting balanced accuracy of 0.825, and a fair balance between sensitivity (90%) and specificity (74%). These models consistently identified a set of psychological traits (optimism, perceived ability to cope with trauma, resilience as trait, ability to comprehend the illness), and subjective perceptions of personal functionality (physical, social, cognitive) as key factors accounting for concurrent depression symptoms. A comprehensive supervised learning pipeline is proposed for the identification of predictors of depression symptoms which could severely impede adaptation to illness.