Evaluating Surface-Enhanced Raman Spectroscopy as a Potential Blood-Based Tool for Crohn’s Disease Detection

In a recent study published in the International Journal of Molecular Sciences, a team of researchers in Romania have reported that serum-based surface-enhanced Raman spectroscopy (SERS), combined with machine learning (ML) analysis, can distinguish patients with Crohn’s disease from healthy individuals with high accuracy, highlighting the technique’s potential as a minimally invasive diagnostic aid for inflammatory bowel disease.¹

What is Crohn’s Disease?

Crohn’s Disease is an inflammatory bowel disease that results in swelling and irritation in the digestive tract.^2,3 It is a lifelong condition with no known cure, but there are therapies that exist that help manage symptoms and improve a patient’s quality of life.^2,3 However, this means that early diagnosis and detection is important, which means reducing the amount of time it takes to diagnose the disease.

The issue, of course, is that current diagnostic methods are time-consuming. Methods such as endoscopy, imaging studies, laboratory testing, and histopathological assessment are also costly and invasive.¹ As a result, researchers have been investigating alternative approaches capable of identifying disease-related biochemical changes through less invasive procedures.

What did the researchers do in their study?

In their study, researchers analyzed serum samples from age- and sex-matched patients with Crohn’s disease and healthy control subjects. Of the 54 participants initially enrolled, 51 participant-level spectra met quality requirements and were retained for final classification analyses.¹

The team employed SERS, which is a technique that amplifies Raman scattering signals to improve the detection of molecular information present in biological samples. Spectral data were processed using Python-based analytical workflows and subjected to two supervised ML classification approaches: principal component analysis-linear discriminant analysis (PCA-LDA) and partial least squares-discriminant analysis (PLS-DA).¹

Despite the small sample size, the researchers took steps to make sure the model evaluation was robust. By using a leave-one-out cross-validation protocol, all preprocessing, dimensionality reduction, and model training steps were performed within each validation cycle to minimize the risk of information leakage.¹

How did both classification models perform?

Both classification models successfully differentiated Crohn’s disease patients from healthy controls. The PCA-LDA model achieved a sensitivity of 85.19%, specificity of 91.67%, accuracy of 88.24%, and an area under the receiver operating characteristic curve (AUC) of 0.881.¹

The PLS-DA model performed slightly better. This model achieved 88.89% sensitivity, 95.83% specificity, 92.16% accuracy, and an AUC of 0.937.¹ These results suggest that serum SERS data contain disease-specific biochemical information that can be leveraged for classification purposes.¹

Beyond diagnostic performance, the study also identified several spectral regions associated with Crohn’s disease. Using Variable Importance in Projection (VIP) analysis, the researchers highlighted discriminatory vibrational bands at 498, 639, 728, 813, 1136, 1205, 1443, 1579, and 1657 cm⁻¹.¹ Statistical testing confirmed significant differences between patient and control groups at these wavelengths.

According to the authors, these spectral features may reflect alterations in purine metabolism, protein conformation, lipid composition, and nucleic acid-related molecular signals.¹ Such findings align with growing evidence that Crohn’s disease is associated with systemic biochemical changes that extend beyond localized intestinal inflammation.¹

What were the limitations of this study?

The findings of this study should not be interpreted as gospel. There were some limitations, including the small sample population and pilot-scale design. While the findings did reveal important insights about the feasibility of combining serum SERS measurements with multivariate ML analysis to identify Crohn’s disease-associated biochemical patterns, additional research will be needed to determine whether the identified spectral biomarkers remain consistent across larger and more diverse patient groups.¹

Future investigations are expected to examine whether SERS-derived signatures can distinguish Crohn’s disease from other gastrointestinal and inflammatory conditions, including ulcerative colitis, infectious colitis, irritable bowel syndrome, and autoimmune diseases.¹ Longitudinal studies may also help determine whether spectral changes correlate with disease progression, treatment response, or remission status.¹

References

1. Valean, D.; Zaharie, R.; Toma, V.; Onaciu, A.; Borsa, R.-M.; Stiufiuc, R.-I.; Fetti, A.; Dohi, B.; Popa, C.; et al. Exploratory Serum-Based Surface-Enhanced Raman Spectroscopy Analysis in Crohn’s Disease: A Pilot Cross-Sectional Study. Int. J. Mol. Sci. 2026, 27 (12), 5180. DOI: 10.3390/ijms27125180
2. Mayo Clinic, Crohn's Disease. Mayo Clinic. Available at: https://www.mayoclinic.org/diseases-conditions/crohns-disease/symptoms-causes/syc-20353304 (Accessed June 23^rd, 2026).
3. Cleveland Clinic, Crohn's Disease. Cleveland Clinic. Available at: https://my.clevelandclinic.org/health/diseases/9357-crohns-disease (Accessed June 23^rd, 2026).