Abstract
Background: Melanoma, a highly aggressive form of skin cancer, necessitates early detection to significantly improve survival rates. Traditional diagnostic techniques, such as white-light imaging (WLI), are effective but often struggle to differentiate between melanoma subtypes in their early stages. Methods: The emergence of the Spectrum-Aided Vison Enhancer (SAVE) offers a promising alternative by utilizing specific wavelength bands to enhance visual contrast in melanoma lesions. This technique facilitates greater differentiation between malignant and benign tissues, particularly in challenging cases. In this study, the efficacy of the SAVE is evaluated in detecting melanoma subtypes including acral lentiginous melanoma (ALM), melanoma in situ (MIS), nodular melanoma (NM), and superficial spreading melanoma (SSM) compared to WLI. Results: The findings demonstrated that the SAVE consistently outperforms WLI across various key metrics, including precision, recall, F1-scorw, and mAP, making it a more reliable tool for early melanoma detection using the four different machine learning methods YOLOv10, Faster RCNN, Scaled YOLOv4, and YOLOv7. Conclusions: The ability of the SAVE to capture subtle spectral differences offers clinicians a new avenue for improving diagnostic accuracy and patient outcomes.