Possible integration of artificial intelligence with photodynamic therapy and diagnosis: A review

Cancer remains a deadly disease with a low median survival rate. The increase in cancer mortality rates is attributed to limitations in diagnosis, prognosis prediction, and therapeutic interventions. Photodynamic diagnosis (PDD) and photodynamic therapy (PDT) are promising light-based systems used for the diagnosis and treatment of cancer. They incorporate photosensitisers (PSs), light and oxygen to visualize or eradicate cancer cells. Even though nanotechnology has advanced these techniques, several limitations, such as target selectivity, PS dose, prognosis, poor light delivery, and inaccurate diagnosis, have hampered their overall efficiency. Artificial intelligence (AI) is a branch of computer science that focusses on predictions and automation, playing a pivotal role in expediting drug discovery and promoting precision in healthcare. AI algorithms have the potential to optimise the design of PSs for improved tissue penetration and affinity, enhancing target-selectivity via identification of novel biomarkers and therapeutic targets, and optimising light delivery parameters for uniform light propagation in tissues. Therefore, this review highlights advancements in the integration of AI in PDD and PDT applications over the last decade, as well as a new perspective on how AI technology can improve PDD and PDT and continue to improve human health in the future.