Multi-algorithmic Genetic Disease Detection Using Pupillometry

Inherited retinal diseases (IRDs) present a significant challenge in paediatric diagnosis due to their diverse genetic and clinical manifestations. The proposed approach integrates a dedicated pupillometer with a custom machine learning system, employing algorithms such as SVM, KNN, Random Forest, Gradient Boosting, LSTM, and BiLSTM . The system addresses the limitations of invasive tests, offering a non-intrusive and child-friendly alternative. Through comprehensive evaluation, the Clinical Decision Support System (CDSS) demonstrates promising diagnostic accuracy, showcasing its potential as a valuable tool in pediatric ophthalmology. This holistic approach represents a significant step towards more accessible and reliable early diagnosis of IRDs, paving the way for timely interventions to mitigate severe visual deficits in affected children.