Mutation Endmember Library Sparse Mixed Abundance Estimation Model for Glioma Margin Determination with Raman...

Area of Science:

• Neuro-oncology
• Spectroscopy
• Computational Biology

Background:

• The glioma margin, a transitional zone between tumor and normal brain tissue, is difficult to delineate using conventional methods.
• Accurate identification of the tumor boundary is crucial for effective glioblastoma surgery.

Purpose of the Study:

• To develop and validate a novel abundance estimation method for Raman spectral unmixing of glioma marginal tissues.
• To enable accurate and real-time determination of the tumor surgical boundary during operations.

Main Methods:

• Application of abundance estimation to Raman spectral unmixing of glioma marginal tissues.
• Introduction of a mutation endmember library sparse mixed abundance estimation model.
• Utilizing group sparse endmember bundle decomposition and fractionally mixed norms for enhanced spectral analysis.

Main Results:

• The novel method demonstrated superior accuracy, stability, and generalization ability compared to conventional techniques in analyzing 112 human glioma margin tissues.
• The approach effectively accounts for spectral variations within glioma tissues.
• Accurate real-time determination of the tumor surgical boundary was achieved.

Conclusions:

• Miniature Raman spectroscopy holds significant potential for in vivo, noninvasive intraoperative margin assessment.
• The developed abundance estimation model offers a robust solution for identifying fuzzy and uncertain tissue boundaries.
• This advancement can lead to improved surgical outcomes for glioblastoma patients.