Research on colorectal cancer detection based on Raman spectroscopy and machine learning

Colorectal cancer is one of the most common malignant tumors in humans, with increasing incidence and mortality rates. Early detection of lesions is crucial for improving patients' survival probability and quality of life. Machine learning algorithms play an increasingly important role in medical detection and diagnosis, especially in tumor classification. This study performed Raman spectroscopy analysis on 56 intestinal tissue sections (29 lesions and 27 healthy samples). Based on the experimental results of Raman spectroscopy, the performance characteristics of various machine learning classifiers in colorectal cancer classification were summarized: Linear Discriminant Analysis (LDA) and Extreme Learning Machine (ELM) showed excellent performance and high training efficiency; Support Vector Machine (SVM) had moderate performance with high stability but longer training time; while Artificial Neural Network (ANN) provided good specificity indicators, its sensitivity was less than ideal. The research on machine learning classifiers provides a new direction for Raman spectroscopy-based detection of colorectal cancer.