Artificial Intelligence in Advancing Optical Coherence Tomography for Disease Detection and Cancer Diagnosis: A scoping review
A partir d'une revue de la littérature, cet article examine comment l'intelligence artificielle peut améliorer les performances cliniques de la tomographie par cohérence optique pour diagnostiquer certains cancers et autres maladies
Background : OCT (Optical Coherence Tomography) functions as a high-resolution non-invasive imaging technology that serves multiple applications within ophthalmology and cardiology and dermatology as well as oncology. The adoption of OCT technology showed major diagnostic progress but medical professionals still face obstacles in complex picture interpretation as well as inconsistent accuracy rates. The implementation of Artificial Intelligence systems that use machine learning (ML) and deep learning (DL) functions has enabled OCT to analyze images automatically while offering better diagnostic precision.
Methods : The study investigates medical sector applications of OCT technology and scrutinizes how Artificial Intelligence facilitates improved clinical performance of OCT. Research conducted on peer-reviewed studies analyzed how AI improves OCT technology by enabling automatic disease detection and real-time image modification and clinical support functions.
Results : The medical field underwent a revolutionary change due to OCT technology that enables improved detection of diseases alongside better patient healing outcomes for retinal conditions and cardiovascular problems and epithelial cancer cases. Real-time surgical oncology decision-making occurs through AI by improving both OCT classification of diseases and detection of tumor margins simultaneously. Convolutional neural networks under artificial intelligence control exhibit strong ability to detect normal and abnormal tissues thus enabling earlier cancer detection with improved medical treatment accuracy. Solving active problems with two main issues stands as the key requirement for progress as clinicians work with uncertain model validity and incomplete dataset similarities in clinical settings.
Conclusions : Clinical procedures benefit from important improvements when OCT networks integrate AI command systems in their operations. Future research demands model optimization for AI technology together with the solution of dataset biases and better implementation in medical settings. AI integration with OCT technology points to substantial potential for healthcare detection developments as well as treatment solutions for individual patients in cancer medicine.
European Journal of Surgical Oncology , résumé, 2025