- Received December 03, 2024
- Accepted April 14, 2025
- Publication May 31, 2025
- Visibility 8 Views
- Downloads 4 Downloads
- DOI 10.18231/j.ijooo.2025.005
-
CrossMark
- Citation
Advancements in artificial intelligence for ocular oncology: Enhancing diagnostic accuracy and prognostic capabilities in eye tumors
- Author Details:
-
Ankit Sanjay Varshney *
-
Debanjali Bhattacharjee
Ankit Sanjay Varshney *
Background: Ocular tumors, including uveal melanoma, retinoblastoma, and other rare neoplasms, present significant diagnostic and therapeutic challenges. Early and accurate diagnosis is vital for optimizing outcomes and improving survival rates.
Objective: This narrative review explores the applications of artificial intelligence (AI) in ocular oncology, emphasizing its role in enhancing diagnostic accuracy, risk stratification, and personalized treatment planning.
Materials and Methods: A systematic literature search was conducted across databases such as PubMed, Scopus, and Embase to identify studies published from 2010 to 2024. Key AI techniques, including convolutional neural networks (CNNs), deep learning (DL), and radiomics, were examined, along with their integration into imaging modalities such as optical coherence tomography (OCT) and fundus photography.
Results: AI tools demonstrated high accuracy in distinguishing uveal melanoma from benign nevi (87.6%) and early-stage retinoblastoma detection with sensitivity and specificity rates over 90%. Radiomics facilitated risk stratification by extracting quantitative tumor features. Hyperspectral imaging combined with AI showed promise in detecting less common ocular tumors. However, ethical concerns, data heterogeneity, and a lack of standardized imaging protocols pose barriers to clinical adoption.
Conclusion: AI holds transformative potential for ocular oncology, offering accurate, efficient, and personalized diagnostic solutions. Addressing challenges such as data quality, ethical compliance, and model generalizability through interdisciplinary collaboration is essential to fully realize its clinical impact.
Keywords: Artificial intelligence, Ocular oncology, Uveal melanoma, retinoblastoma, Deep learning, Radiomics, Diagnostic imaging, Personalized medicine.