Mitura Przemysław, Paja Wiesław, Młynarczyk Grzegorz, Kowalski Radosław, Bar Krzysztof, Depciuch Joanna
Department of Urology and Oncological Urology, Medical University of Lublin, Jaczewskiego 8, 20-954 Lublin, Poland.
Department of Artificial Intelligence, Institute of Computer Science, University of Rzeszow, Pigonia 1, 35-310 Rzeszów, Poland.
Spectrochim Acta A Mol Biomol Spectrosc. 2026 Jan 5;344(Pt 1):126661. doi: 10.1016/j.saa.2025.126661. Epub 2025 Jul 10.
This study investigates the potential of Raman spectroscopy in distinguishing between healthy individuals and prostate cancer patients using urine samples. The Boruta algorithm was applied to Raman spectral data in two distinct wavenumber regions: 800-1800 cm (fingerprint region) and 2800-3000 cm (lipid region). The algorithm identified important spectral features from both regions that were used to construct decision trees for classification. Key wavenumbers in the fingerprint region (1009 cm) and high-wavenumber region (2937 cm) were found to be significant markers for prostate cancer detection. Principal Component Analysis (PCA) revealed that the intensity of these markers effectively separated healthy and cancerous samples, with the 1009 cm marker showing higher discriminative power. Furthermore, four classification models: Decision Tree (DT), k-Nearest Neighbors (kNN), Random Forest (RF), and Support Vector Machine (SVM) were evaluated for their performance in classifying urine samples based on Raman spectral features. The RF and kNN models exhibited the best overall performance, with high accuracy and sensitivity, particularly in the 800-1800 cm region. The study also explored the correlation between Raman markers and clinical parameters, finding that the 2937 cm marker had strong correlations with critical clinical variables like Gleason scores and MRI PIRADS scores, suggesting its utility for prostate cancer diagnosis and staging. These findings highlight the potential of Raman spectroscopy as a non-invasive tool for prostate cancer detection and monitoring.
本研究调查了拉曼光谱法利用尿液样本区分健康个体和前列腺癌患者的潜力。将博鲁塔算法应用于两个不同波数区域的拉曼光谱数据：800 - 1800厘米（指纹区）和2800 - 3000厘米（脂质区）。该算法从这两个区域识别出重要的光谱特征，用于构建分类决策树。发现指纹区（1009厘米）和高波数区（2937厘米）的关键波数是前列腺癌检测的重要标志物。主成分分析（PCA）表明，这些标志物的强度有效地分离了健康样本和癌性样本，1009厘米的标志物显示出更高的判别能力。此外，评估了四种分类模型：决策树（DT）、k近邻（kNN）、随机森林（RF）和支持向量机（SVM）基于拉曼光谱特征对尿液样本进行分类的性能。RF和kNN模型表现出最佳的整体性能，具有高准确性和敏感性，特别是在800 - 1800厘米区域。该研究还探讨了拉曼标志物与临床参数之间的相关性，发现2937厘米的标志物与 Gleason评分和MRI PIRADS评分等关键临床变量有很强的相关性，表明其在前列腺癌诊断和分期中的效用。这些发现突出了拉曼光谱法作为一种用于前列腺癌检测和监测的非侵入性工具的潜力。
