Machine learning analysis of serum ATR-FTIR spectra for colorectal cancer discrimination

Colorectal cancer (CRC) remains a major cause of cancer-related mortality, highlighting the need for complementary minimally invasive approaches able to support early CRC detection. This study investigated serum attenuated total reflection Fourier-transform infrared (ATR-FTIR) spectroscopy combined with machine learning analysis for discriminating CRC cases from non-cancer controls. ATR-FTIR spectra were acquired from dried serum samples of 100 CRC cases and 100 non-cancer controls and analyzed in the fingerprint and high-wavenumber spectral regions. Second-derivative spectra were used to enhance spectral resolution and to perform variable importance in projection analysis, which guided the selection of discriminant spectral regions before classification. The most relevant variables were mainly assigned to protein-related vibrations, phosphate-containing biomolecules, and carbohydrate-associated molecular structures. A PCA–ANN classifier was developed to reduce spectral dimensionality while preserving model parsimony. In internal validation, the model achieved 87.5% accuracy, 90.0% sensitivity, 85.0% specificity, and an AUC of 0.941. Preliminary evaluation on an independent external dataset of 28 subjects yielded 85.7% accuracy and an AUC of 0.920. These findings support the feasibility of serum ATR-FTIR fingerprints combined with interpretable machine learning for CRC/control discrimination, although larger prospective and demographically balanced external cohorts are required to assess its potential utility for CRC early detection.

Authors

Institutions

Publication Details

Journal
Scientific Reports
Published
2026-11-21
DOI
https://doi.org/10.1038/s41598-026-81165-x
Primary Topic
Spectroscopy Techniques in Biomedical and Chemical Research
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Machine learning analysis of serum ATR-FTIR spectra for colorectal cancer discrimination

Vitoantonio Bevilacqua, Antonio Brunetti, Andrea Zifarelli, G. Giannelli et al.
Scientific Reports
Spectroscopy Techniques in Biomedical and Chemical Research
article

Machine learning analysis of serum ATR-FTIR spectra for colorectal cancer discrimination

Vitoantonio Bevilacqua, Antonio Brunetti, Andrea Zifarelli, G. Giannelli, A. Picciariello, D. Spagnolo, M. Giglio, L. Vincenti, A. Ancona, I. Grassi, P. L. Pesole, R. Donghia
article en

Abstract

Colorectal cancer (CRC) remains a major cause of cancer-related mortality, highlighting the need for complementary minimally invasive approaches able to support early CRC detection. This study investigated serum attenuated total reflection Fourier-transform infrared (ATR-FTIR) spectroscopy combined with machine learning analysis for discriminating CRC cases from non-cancer controls. ATR-FTIR spectra were acquired from dried serum samples of 100 CRC cases and 100 non-cancer controls and analyzed in the fingerprint and high-wavenumber spectral regions. Second-derivative spectra were used to enhance spectral resolution and to perform variable importance in projection analysis, which guided the selection of discriminant spectral regions before classification. The most relevant variables were mainly assigned to protein-related vibrations, phosphate-containing biomolecules, and carbohydrate-associated molecular structures. A PCA–ANN classifier was developed to reduce spectral dimensionality while preserving model parsimony. In internal validation, the model achieved 87.5% accuracy, 90.0% sensitivity, 85.0% specificity, and an AUC of 0.941. Preliminary evaluation on an independent external dataset of 28 subjects yielded 85.7% accuracy and an AUC of 0.920. These findings support the feasibility of serum ATR-FTIR fingerprints combined with interpretable machine learning for CRC/control discrimination, although larger prospective and demographically balanced external cohorts are required to assess its potential utility for CRC early detection.

Scientific Reports
University of Salento (IT), Gastroenterology Hospital "Saverio de Bellis" (IT), Polytechnic University of Bari (IT)
Spectroscopy Techniques in Biomedical and Chemical Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon