Malignant Tumours

The goal of this observational study is to explore whether a Raman-based, deep learning-assisted approach can be used to develop an effective method for early pan-cancer screening. The study includes healthy individuals, patients at risk of cancer, and patients with diagnosed cancers. The main questions it aims to answer are: * Evaluating the deep-learning model's accuracy and specificity in identifying cancer-specific features in Raman spectral data and determining whether this method can accurately classify patients based on risk. * Identifying which model is more adaptable to the Raman spectrum * Providing an interpretable analysis of the model-generated diagnosis Participants are already being diagnosed and follow-up to determine the type of cancer.