Emerging light-based strategies in cancer theranostics: Photodynamic therapy, nanomedicine, and precision oncology

The integration of artificial intelligence (AI) into medical treatments is reshaping precision and personalisation in phototherapy (PT) and photodynamic therapy (PDT). PT, commonly used for conditions like psoriasis and vitiligo, uses light to modulate cellular responses, but traditional approaches often lack precision. PDT, which combines light with a photosensitising agent to destroy targeted cells, is widely applied in oncology. Both therapies benefit significantly from AI-enhanced targeting and monitoring to improve outcomes and minimise side effects. AI, particularly deep learning (DL) models has shown high accuracy in tumour imaging often exceeding 95% in various diagnostic tasks. In general, AI-assisted optical imaging achieves ∼90-95% accuracy for tumour detection and ∼85-95% accuracy for margin delineation/segmentation, depending on the imaging modality and dataset size. These tools, when integrated with optical devices, enable real-time treatment monitoring and improved decision-making during therapy. Beyond imaging, AI frameworks are increasingly being explored to support adaptive treatment planning and response assessment across light-based therapeutic platforms. AI shows potential to further support personalised therapy by analysing patient histories, genetics, and treatment responses to recommend optimal drug and light dosages. Moreover, AI can accelerate drug discovery by predicting the efficacy of new photosensitisers and optimising light delivery for consistent treatment. Clinical decision support systems powered by AI help tailor PDT strategies based on patient data. Though still in early stages, AI's role in PT and PDT is expanding, with future directions focusing on wearable technologies, advanced algorithms, and integrated data platforms to further enhance therapeutic precision and patient care.