Aug 2026· Karnataka Journal of Surgery· Vol 3, pp. 61-75· 0 citations· 18 references
TL;DR
This comprehensive review delves into the fundamental concepts of radiomics and AI, summarises their current applications in BC imaging, and explores their evolving roles in clinical practice, highlighting recent advances and presenting case studies demonstrating the clinical impact.
Abstract
Breast cancer (BC) continues to be the most prevalent malignancy affecting women globally, representing a major public health concern, with significant morbidity and mortality. Early detection, accurate diagnosis, and precise characterisation of breast lesions are important to improve patient outcomes and survival rates. Conventional imaging modalities, such as mammography, ultrasound, and magnetic resonance imaging (MRI), have played pivotal roles in BC diagnosis but face limitations related to subjective interpretation, variability between radiologists, and challenges in detecting biologically aggressive subtypes. Radiomics and artificial intelligence (AI) have emerged as revolutionary adjuncts to enhance the diagnostic and prognostic capabilities of breast imaging. Radiomics involves the extraction of high-dimensional quantitative imaging features from standard medical images that are imperceptible to the human eye. These features can reveal tumour heterogeneity, microenvironment characteristics, and biological behaviour, thereby enriching information traditionally derived from visual inspection. AI, particularly through machine learning and deep learning models, enables automated analysis, pattern recognition, and prediction of clinical outcomes with high accuracy and reproducibility. The integration of radiomics and AI into BC imaging workflows holds the potential to shift the paradigm towards precision oncology, offering individualised risk stratification, early prediction of treatment response, and real-time decision support. However, this field faces significant challenges, including issues related to data standardisation, reproducibility, model validation, regulatory approval, and clinical integration. Ethical considerations regarding the data privacy, bias, and explainability of AI algorithms also remain critical hurdles. This comprehensive review delves into the fundamental concepts of radiomics and AI, summarises their current applications in BC imaging, and explores their evolving roles in clinical practice. It highlights recent advances, presents case studies demonstrating the clinical impact, and discusses ongoing research efforts aimed at overcoming the existing limitations. Furthermore, future directions, including the integration of radio genomics, explainable AI (XAI), and multi-omics approaches, were thoroughly examined to provide a roadmap for the clinical applicability of these technologies. As the convergence of advanced imaging analytics and computational intelligence continues to mature, radiomics and AI have been poised to redefine BC management, ushering in a new era of more accurate, efficient, and personalised patient care.
Breast cancer represents a leading public health problem due to its high incidence, morbidity, and mortality. Significant geographic differences in disease burden are influenced not only by socioeconomic and behavioural factors but also by the resources of health systems, access to healthcare, and the quality of diagnostic programs. Modern mammography screening programs substantially contribute to early disease detection, while existing challenges include limited resources, restricted access and inequalities in healthcare, high costs, variability in result interpretation, and the need for additional and invasive diagnostic procedures. To address these challenges, artificial intelligence systems have been developed, employing complex algorithms, machine learning, and deep learning for the automated analysis of mammographic images. These systems can improve diagnostic accuracy, reduce reading time, lower false-negative rates, and decrease the need for repeat diagnostic procedures. Research has demonstrated that the accuracy achieved through artificial intelligence algorithms is comparable to that of radiologists, and optimal results may be obtained by combining radiologist's assessments with artificial intelligence methods, particularly in the role of the first reader. Key limitations of these methods include insufficient adaptability to diverse populations and the risk of over-reliance on algorithm results by less experienced radiologists. Beyond technical challenges, an important public health aspect concerns women's attitudes toward the use of artificial intelligence, as the level of trust in this technology may affect screening participation rates. The integration of artificial intelligence into screening programs requires careful evaluation of benefits and limitations, transparency in technology use, and assurance of human oversight. Proper implementation and public education can contribute to improved early diagnosis, greater screening accuracy, and more efficient use of health system resources.
Kristina Stamenković, V. Mijatović-Jovanović, D. Milijašević· Glasnik javnog zdravlja· 0 citations
Background: Breast cancer remains a major public health burden, and improvements in imaging-based detection, risk stratification, and diagnostic workflow may support earlier diagnosis, reduce false-negative interpretations, and improve population-level screening efficiency. Recent advances in quantitative breast imaging, radiomics, and artificial intelligence (AI) have expanded the role of imaging beyond lesion detection toward tumor characterization, prognostication, and precision medicine. However, evidence regarding their diagnostic and predictive performance remains heterogeneous. This study aimed to systematically evaluate the diagnostic accuracy of quantitative imaging biomarkers, radiomics, and AI-based approaches in breast cancer and to assess their associations with tumor biological characteristics.
Methods: A systematic review and meta-analysis were conducted in accordance with the PRISMA 2020 guidelines. MEDLINE (PubMed), Embase, Scopus, Web of Science, Cochrane CENTRAL, and IEEE Xplore were searched for studies published between January 2015 and December 2024. Eligible studies evaluated quantitative imaging biomarkers, radiomics, or AI applications in breast cancer diagnosis or characterization. Diagnostic performance measures, including sensitivity, specificity, area under the curve (AUC), and biomarker associations with pathological features, were extracted. A random-effects meta-analysis was performed to pool AUC values where appropriate.
Results: Thirteen studies involving quantitative imaging biomarkers (n = 5), radiomics (n = 4), and AI-based detection systems (n = 4) met the inclusion criteria. For differentiation of benign and malignant breast lesions, apparent diffusion coefficient (ADC) measurements demonstrated excellent diagnostic performance, with a pooled AUC of 0.94 (95% CI: 0.91-0.97). Individual studies reported sensitivities ranging from 84.1% to 92.5% and specificities from 90.2% to 91.1%, with optimal ADC thresholds between 1.23 and 1.30 × 10−3 mm2/s. Across studies evaluating tumor grade, lower ADC values were generally associated with higher-grade tumors; however, quantitative pooling was not retained because of the small number of contributing studies, non-independent comparisons, and substantial heterogeneity. Radiomics studies achieved molecular subtype classification accuracies ranging from 77% to 100%, particularly when derived from magnetic resonance imaging (MRI) and ADC maps. AI systems showed excellent diagnostic performance, with AUCs ranging from 0.876 to 0.97, consistently matching or exceeding radiologist performance. Hybrid AI-radiologist approaches yielded the highest diagnostic accuracy and reduced false-negative interpretations.
Conclusion: Quantitative imaging biomarkers, particularly ADC, provide useful diagnostic information and may offer prognostic insights in breast cancer. Radiomics shows strong potential for noninvasive molecular characterization, while AI-based systems achieve high diagnostic accuracy and improve clinical workflow when integrated with radiologist interpretation. These technologies represent promising components of precision breast imaging.
Dhekra Mugahed, Marwah Alharbi, Alanoud E. Dalak et al.· Saudi Journal of Public Heal...· 0 citations
Prostate cancer (PCa) is the most prevalent malignant tumor in the urogenital system among men worldwide. Due to its subtle early symptoms and strong tumor heterogeneity, traditional diagnostic methods relying on a single prostate-specific antigen (PSA) initial screening and subjective imaging evaluations often lead to high false positives, overt biopsies, and missed small lesions. The rapid development of artificial intelligence (AI) provides innovative solutions to overcome these clinical bottlenecks. This article comprehensively reviews the application of AI in the early intelligent diagnosis of PCa. In the fields of ultrasound, magnetic resonance imaging (MRI), and positron emission tomography/computed tomography (PET/CT) imaging, AI significantly enhances the accuracy of target lesion identification. It achieves this by deep decoding high-dimensional quantitative features and effectively reducing subjective bias. In non-invasive liquid biopsy, AI-driven multi-omics networks have successfully addressed challenging screening blind spots, such as the PSA gray zone. In light of current challenges such as limited model generalization capability and the “black box effect” of algorithms, this article looks forward to the development prospects of constructing multimodal fusion models based on federated learning and explainable AI (XAI), aiming to promote the transition of PCa diagnosis and treatment from algorithm development to real clinical decision support.
Jun Xie, Ziwei Yin, Ruisong Gao et al.· Frontiers in Oncology· 0 citations
Breast cancer remains one of the most prevalent malignancies worldwide, where early diagnosis significantly improves survival rates and treatment outcomes. Recent advances in artificial intelligence (AI) and deep learning have demonstrated substantial potential to enhance histopathological image analysis and support precision oncology. This study presents an extended deep learning framework for early breast cancer detection using histopathological images and explores its potential application in personalized treatment strategies. The publicly available IDC_regular dataset comprising 277,524 image patches extracted from 162 whole-slide breast cancer specimens was utilized. A Convolutional Neural Network (CNN)-based architecture was employed for feature extraction and binary classification of invasive ductal carcinoma (IDC) positive and negative cases. The proposed framework incorporated image pre-processing, OpenCV-based transformations, data normalization, model training, and evaluation using precision, recall, F1-score, and accuracy metrics. Experimental results achieved an overall classification accuracy of 92%, with precision and recall values demonstrating reliable detection performance. Furthermore, saliency mapping techniques were introduced to improve interpretability and localize diagnostically relevant regions. The extracted deep features provide a foundation for future multi-class tumour characterization, risk stratification, and treatment response prediction. The findings suggest that AI-assisted pathology can reduce diagnostic workload, improve detection efficiency, and support personalized clinical decision-making. Future work will focus on integrating clinical, genomic, and treatment datasets to develop comprehensive precision oncology systems.
Afroja Nahida· Journal of Medical Clinical...· 0 citations
The continued integration of AI has the potential to improve diagnostic accuracy, streamline radiological workflows, and support more personalized patient care in the next generation of diagnostic imaging.
Mr. Vishal Walia, Mr. Honey Thakur, Ms. Ashwarya Sharma et al.· PAIN, JOINTS, SPINE· 0 citations
Breast cancer is a heterogeneous disease, and accurate preoperative identification of molecular subtypes is essential for guiding individualized treatment and prognostic evaluation. However, histopathological assessment, the current gold standard, is invasive and limited by intratumoral heterogeneity, underscoring the need for reliable noninvasive alternatives. Radiomics based on dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) has shown promise for molecular subtype prediction; nevertheless, the optimal radiomic features for automated molecular classification remain to be fully elucidated. To address this gap, this study aims to identify optimal DCE-MRI radiomic features and to develop radiomics-based machine learning models for predicting breast cancer molecular subtypes. Specifically, regions of interest were manually delineated on DCE-MRI images for feature extraction, followed by feature selection to identify the most informative radiomic parameters. Four machine learning classifiers (RF, SVM, LR, GBDT) were then constructed to predict five molecular subtypes of breast cancer. The experimental results demonstrate that the optimized DCE-MRI–based radiomics model effectively predicts breast cancer molecular subtypes. In particular, the gradient boosting decision tree (GBDT) model combined with rigorous feature selection shows high predictive performance, highlighting its strong potential for noninvasive, accurate, and clinically applicable molecular classification of breast cancer.
Yang Lou, Liang Zhao· Frontiers in Oncology· 0 citations
