Utilizing Raman spectroscopy for urinalysis to diagnose acute kidney injury stages in cardiac surgery patients

The successful treatment and improvement of acute kidney injury (AKI) depend on early-stage diagnosis. However, no study has differentiated between the three stages of AKI and non-AKI patients following heart surgery. This study will fill this gap in the literature and help to improve kidney disease management in the future. In this study, we applied Raman spectroscopy (RS) to uncover unique urine biomarkers distinguishing heart surgery patients with and without AKI. Given the amplified risk of renal complications post-cardiac surgery, this approach is of paramount importance. Further, we employed the partial least squares-support vector machine (PLS-SVM) model to distinguish between all three stages of AKI and non-AKI patients. We noted significant metabolic disparities among the groups. Each AKI stage presented a distinct metabolic profile: stage 1 had elevated uric acid and reduced creatinine levels; stage 2 demonstrated increased tryptophan and nitrogenous compounds with diminished uric acid; stage 3 displayed the highest neopterin and the lowest creatinine levels. We utilized the PLS-SVM model for discriminant analysis, achieving over 90% identification rate in distinguishing AKI patients, encompassing all stages, from non-AKI subjects. This study characterizes the incidence and risk factors for AKI after cardiac surgery. The unique spectral information garnered from this study can also pave the way for developing an in vivo RS method to detect and monitor AKI effectively.急性肾损伤（acute kidney injury, AKI）的成功治疗与病情改善有赖于早期诊断，但目前尚无研究对心脏术后急性肾损伤患者的三个分期与非急性肾损伤患者进行区分。本研究将填补这一文献空白，助力未来肾脏疾病管理水平的提升。本研究采用拉曼光谱（Raman spectroscopy, RS）技术，挖掘可区分心脏术后合并与未合并急性肾损伤患者的独特尿液生物标志物。鉴于心脏术后肾脏并发症风险显著升高，该方法具有极高的应用价值。进一步，本研究采用偏最小二乘-支持向量机（partial least squares-support vector machine, PLS-SVM）模型，对急性肾损伤三个分期患者与非急性肾损伤患者进行区分。研究发现各组间存在显著代谢差异：各急性肾损伤分期均具有独特的代谢特征：1期患者血尿酸水平升高、肌酐水平降低；2期患者色氨酸与含氮化合物水平升高，血尿酸水平降低；3期患者新蝶呤水平最高且肌酐水平最低。本研究采用PLS-SVM模型开展判别分析，在区分各分期急性肾损伤患者与非急性肾损伤受试者时，识别准确率超过90%。本研究明确了心脏术后急性肾损伤的发病情况与危险因素。本研究获取的独特光谱信息，也可为开发可有效检测与监测急性肾损伤的体内拉曼光谱技术奠定基础。