Data fusion in metabolomic cancer diagnostics

Area of Science:

• Biomedical diagnostics
• Cancer research
• Analytical chemistry

Background:

• Early detection of colorectal cancer (CRC) is crucial for improving patient outcomes.
• Plasma fluorescence spectroscopy has shown potential for non-invasive CRC screening.

Purpose of the Study:

• To enhance the early detection of colorectal cancer by integrating multiple diagnostic modalities.
• To evaluate the combined diagnostic performance of fluorescence spectroscopy, specific biomarkers, and metabolomic profiling.

Main Methods:

• Utilized plasma samples for fluorescence spectroscopy analysis.
• Incorporated established biomarkers: Carcinoembryonic antigen (CEA) and Tissue inhibitor of metalloproteinases-1 (TIMP-1).
• Applied proton nuclear magnetic resonance ((1)H NMR) spectroscopy for metabolomic profiling.

Main Results:

• The integrated approach demonstrated superior performance over single marker detection.
• Combining fluorescence data with CEA, TIMP-1, and (1)H NMR provided a more robust diagnostic profile.
• The multi-modal analysis significantly improved the accuracy for early colorectal cancer detection.

Conclusions:

• An extensive diagnostic profile combining fluorescence spectroscopy, biomarkers (CEA, TIMP-1), and metabolomics ((1)H NMR) is more effective for early colorectal cancer detection.
• This integrated strategy offers a promising advancement in non-invasive cancer diagnostics.
• Future research should focus on validating this multi-modal approach in larger clinical cohorts.