目前普遍应用于前列腺癌早期诊断的前列腺特异性抗原筛查由于其有限的诊断特异性受到广泛质疑。申请者的前期工作发现前列腺癌细胞内脂滴中胆固醇酯的异常积累，并证明它有潜力成为前列腺癌的标志物。但是将胆固醇酯的检测应用于前列腺癌的早期诊断还有很多问题有待解决。首先，针对组织标本的检测只能辅助穿刺活检病理分析，对临床早期诊断的价值有限。其次，目前的检测方法只能提供人为挑选的特定脂滴的成分信息，忽视了不同癌细胞之间的差异性，具有很强的主观性，而且无法实现自动化数据采集和处理。针对以上问题，申请者提出利用高光谱受激拉曼散射显微术结合多元曲线分辨数据处理的方法对前列腺癌尿液标本中胆固醇酯的含量进行定量分析，从而为前列腺癌的早期诊断提供客观准确的检测方法。为证实该假说，本研究将首先设计并实现该成像系统；然后确定针对尿液标本的诊断阈值、灵敏性和特异性；最后评估该检测方法用于前列腺癌早期诊断的价值。

There has been a big debate about the usefulness of prostate specific antigen screening in prostate cancer diagnosis, due to its low specificity and limited capability to identify aggressive forms of cancer. Thus, a better biomarker is urgently needed for prostate cancer screening and diagnosis. In our preliminary study, we have employed Raman spectromicroscopy, which couples stimulated Raman scattering (SRS) microscopy with spontaneous Raman spectroscopy, and found an unexpected, aberrant accumulation of esterified cholesterol in LDs of prostate cancer. Results of our pilot biological studies further revealed that cholesteryl ester (CE) accumulation can be potentially used as a molecular marker for prostate cancer and its aggressiveness. However, there have been questions to be addressed in order to use CE as a marker for prostate cancer diagnosis in clinic. First, analysis of CE in tissues can only provide additional information for biopsy and histopathology, and so its contribution to early diagnosis is limited. Second, Raman spectromicroscopy can only analyze certain picked LDs. This not only introduces subjective factors but also overlooks cancer cell heterogeneity, which significantly reduces accuracy and reliability of this Raman spectromicroscopy-based method. In this proposed project, we hypothesize that through quantitative mapping of CEs in urine samples, hyperspectral SRS microscopy coupled with multivariate curve resolution data analysis can accurately and objectively diagnose prostate cancer. To test this hypothesis, we will first develop the new version of hyperspectral SRS imaging system, and then determine the diagnostic cutoff, sensitivity, and specificity, and finally evaluate the value of this new method in early diagnosis of prostate cancer.