Shivam Dubey
The current trend in ocular oncology involves new technologies and patient-centered approaches. Converging technologies such as Artificial Intelligence (AI), multi-modal imaging, and precision medicine will augment the traditional methods of diagnosing and treating ocular tumors, including but not limited to uveal melanoma, retinoblastoma, and conjunctival malignancies. AI-enabled systems are being integrated to process diverse data types to provide a more comprehensive diagnostic picture for the treating physician; this includes information from Optical Coherence Tomograms (OCT), fundus photographs, histopathologic specimens, and genomic profile(s). This review documents the emerging concept of AI-based multimodal precision oncology in ocular tumor management through its applications along the entire continuum of care including disease detection and risk stratification, therapeutic decision-making, and post-oncological reconstructive procedures. Convolutional Neural Networks (CNNs), other deep-learning architectures, and transformer technology provide high diagnostic accuracy when combined with radiogenomic and liquid biopsy data. In addition, AI-assisted surgical planning and 3D reconstructed surgical techniques are revolutionizing the field of oculoplasty by providing personalized rehabilitation solutions for patients after tumor removal. Whilethese advances are promising, there are still many challenges to overcome in terms of clinical translation due to data issues (i.e., heterogeneity of datasets) and regulatory issues (e.g., lack of prospective validation). This review summarizes current data, highlights key gaps in research, and outlines areas for improvement in developing an integrated AI ocular oncology system which takes a holistic patient approach. Ultimately, we hope to shift from reactive treatment methods to those that are proactive, personalized, and focused on restoring functionality.