Scholarlink

<p>Globally, colorectal cancer (CRC) continues to rank among the top causes of cancer-related morbidity and death. In the current study, surface enhanced Raman spectroscopy (SERS) combined with machine learning was employed for the identification and characterization of blood serum samples of colorectal cancer patients and healthy individuals for the identification of biochemical changes associated with the development of this disease. Significant SERS spectral features are identified which differentiate normal and cancerous samples. For the SERS analysis, silver nanoparticles (Ag-NPs) are utilized as SERS substrate for the analysis of clinically verified blood serum samples from colorectal cancer patients of different stages and control group. The multivariate statistical techniques including principal component analysis (PCA) and partial least square discriminant analysis (PLS-DA) along with support vector machine (SVM) were employed to classify and characterize the spectral features associated with each stage of colorectal cancer. PLS-DA differentiated the SERS spectral groups of colorectal cancerous samples of different stages and healthy individuals with 100% sensitivity, 90% specificity, and area under curve (AUC) value of 80.10%, while SVM showing AUC values between 0.99 and 1.00 for all categories, reflecting near-perfect classification for healthy, stage-1, and stage-3 cases, with 100% accuracy, stage-4 with 97% and stage-2 with only 82.6% indicating that these models are best fit for differentiation and classification of these SERS spectral features of blood serum samples.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multivariate Statistical Techniques in Surface Enhanced Raman Spectroscopy for Colorectal Cancer Biomarker Identification

  • 期刊论文

  • 发表日期：

    2025年6月12日

查看全文
  • Amna Tahir,
  • Rabeea Razaq,
  • Muhammad Irfan Majeed,
  • Haq Nawaz,
  • Nosheen Rashid,
  • Hanen karamti,
  • Arslan Ali,
  • Arslan Yousaf,
  • Munawar Hussain,
  • Muhammad Ali,
  • Abdul Lateef,
  • Sadia Arshad,
  • Maida Ehsan,
  • Noor Ul Sabah,
  • Muhammad Imran

摘要

Globally, colorectal cancer (CRC) continues to rank among the top causes of cancer-related morbidity and death. In the current study, surface enhanced Raman spectroscopy (SERS) combined with machine learning was employed for the identification and characterization of blood serum samples of colorectal cancer patients and healthy individuals for the identification of biochemical changes associated with the development of this disease. Significant SERS spectral features are identified which differentiate normal and cancerous samples. For the SERS analysis, silver nanoparticles (Ag-NPs) are utilized as SERS substrate for the analysis of clinically verified blood serum samples from colorectal cancer patients of different stages and control group. The multivariate statistical techniques including principal component analysis (PCA) and partial least square discriminant analysis (PLS-DA) along with support vector machine (SVM) were employed to classify and characterize the spectral features associated with each stage of colorectal cancer. PLS-DA differentiated the SERS spectral groups of colorectal cancerous samples of different stages and healthy individuals with 100% sensitivity, 90% specificity, and area under curve (AUC) value of 80.10%, while SVM showing AUC values between 0.99 and 1.00 for all categories, reflecting near-perfect classification for healthy, stage-1, and stage-3 cases, with 100% accuracy, stage-4 with 97% and stage-2 with only 82.6% indicating that these models are best fit for differentiation and classification of these SERS spectral features of blood serum samples.