Multicenter, Prospective Clinical Study of the Serum Raman Spectroscopy Intelligent System for the Diagnosis of Prostate Cancer

At present, the most commonly used clinical screening tool is based on prostate-specific antigen (PSA) examination. Because PSA is a tissue-specific rather than a tumor-specific marker, it has low specificity and sensitivity for prostate cancer. Although these PSA-related diagnostic models (PHI, 4Kscore) have been proved to improve the sensitivity and specificity of the early diagnosis of prostate cancer, they still do not meet the requirements of accurate diagnosis. Therefore, it is extremely important to develop a diagnosis tool with higher specificity, sensitivity and accuracy in the current prostate tumor screening strategy.

Raman spectroscopy (Raman Spectrum, RS) as a non-invasive and high specificity of material molecular detection technology, can be obtained in the molecular level, thus sensitive to detect biological samples tumor metabolism related proteins, nucleic acids, lipids and sugar composition of bio-molecules changes. As scientists pointed out in a literature in "chemical society reviews"in 2020, although SERS technology has shown good diagnostic efficacy in lots of preclinical studies in multiple tumors, it is limited to a generally small sample size and lacks external validation. There for, a clinical study of Raman spectra for tumor diagnosis is needed, which meets the following requirements: 1.An objective, fast and practical application of Raman spectral data processing is needed and deep learning method may be the best classification method; 2. It requires multicenter and large clinical samples to train deep learning diagnostic model, and verify its true efficacy through external data of prospective study.

In our preliminary study,we have collected Raman spectra data from a large cohort of 2899 patients and constructed Raman intelligent diagnostic system based on CNN model. The intelligent diagnostic system achieved accuracy of 83%. In order to obtain the highest level of clinical evidence and truly realize clinical transformation, this prospective, multi-center clinical study is designed to verify the intelligent diagnostic system for early diagnosis of prostate cancer.