Summary
This dissertation addresses the global rise in skin cancer incidence by exploring two key approaches. It investigates the potential of non-invasive 3D hyperspectral imaging (HSI) combined with convolutional neural networks (CNNs) to differentiate between malignant and benign skin lesions. The HSI-CNN demonstrated promising accuracy in distinguishing various types of skin tumors, providing a potentially valuable tool for quick and accurate skin cancer detection. also, it assesses the effectiveness and tolerability of two laser-mediated photodynamic therapy (PDT) methods in treating skin cancer precursors and field cancerization. The results indicate that ablative fractional laser (AFXL)-mediated PDT, combined with artificial daylight, is a promising single-visit treatment option for patients with extensive or severe photodamage. Further research is needed to refine these methods and validate their broader applicability.
