Automatic subtyping of Diffuse Large B-cell Lymphomas (DLBCL): Raman-based genetic and metabolic classification

Area of Science:

• Biomedical Optics
• Molecular Spectroscopy
• Oncology

Background:

• Diffuse Large B-cell Lymphoma (DLBCL) is a common, aggressive blood cancer with significant heterogeneity.
• Accurate classification of DLBCL subtypes is crucial for effective treatment selection but remains challenging.
• Existing diagnostic methods struggle with the molecular complexity and metabolic variability of DLBCL.

Purpose of the Study:

• To evaluate Raman imaging as a diagnostic tool for classifying DLBCL cells.
• To develop and present models for automated DLBCL cell identification and subtype assignment.
• To correlate spectral profiles with biochemical, genetic, and metabolic features of DLBCL subtypes.

Main Methods:

• Utilized Raman imaging to acquire spectral data from DLBCL cells.
• Developed classification models based on Raman spectra for automated identification.
• Assigned DLBCL cells to cell-of-origin (COO) subtypes (ABC/GCB) and comprehensive cluster classification (CCC) subtypes (OxPhos/non-OxPhos).

Main Results:

• Raman imaging provides unique biochemical and metabolic information from DLBCL cells.
• Developed models successfully identified DLBCL cells and assigned them to specific subtypes.
• Distinct spectral profiles were identified for each lymphoma subtype, linked to underlying biological characteristics.

Conclusions:

• Raman imaging demonstrates high diagnostic potential for DLBCL classification.
• Automated spectral analysis enables efficient and accurate identification of DLBCL subtypes.
• This technique offers a novel approach to understanding DLBCL heterogeneity and guiding treatment.