Classifying low-grade and high-grade bladder cancer using label-free serum surface-enhanced Raman spectroscopy and support vector machine. (8th February 2018)

• Record Type: Journal Article
• Authors: Zhang, Yanjiao; Lai, Xiaoping; Zeng, Qiuyao; Li, Linfang; Lin, Lin; Li, Shaoxin; Liu, Zhiming; Su, Chengkang; Qi, Minni; Guo, Zhouyi
• Is Part Of: Laser physics. Volume 28:Number 3(2018:Mar.)
• Publication Date: 2018-02-08
• Subjects: Raman spectroscopy -- bladder cancer -- medical diagnostics -- support vector machine; Lasers -- Periodicals; 621.36605
• DOI: 10.1088/1555-6611/aa9d6d
• Languages: English
• ISSNs: 1054-660X

Abstract

Abstract: This study aims to classify low-grade and high-grade bladder cancer (BC) patients using serum surface-enhanced Raman scattering (SERS) spectra and support vector machine (SVM) algorithms. Serum SERS spectra are acquired from 88 serum samples with silver nanoparticles as the SERS-active substrate. Diagnostic accuracies of 96.4% and 95.4% are obtained when differentiating the serum SERS spectra of all BC patients versus normal subjects and low-grade versus high-grade BC patients, respectively, with optimal SVM classifier models. This study demonstrates that the serum SERS technique combined with SVM has great potential to noninvasively detect and classify high-grade and low-grade BC patients.