# Time‐Resolved Fluorescence and Diffuse Reflectance (TRF‐DR) Spectroscopy for Heterogeneous Breast Tissue Classification and Tumor Margin Assessment

## TL;DR

A new spectroscopy system helps identify breast tumor boundaries during surgery, offering a fast and portable solution with performance similar to current methods.

## Contribution

A TRF-DR spectroscopy system with weighted logistic regression is introduced for tumor margin assessment in heterogeneous breast tissue.

## Key findings

- The TRF-DR system achieved 77% sensitivity and 68% specificity in classifying tumor regions.
- The system's performance is comparable to existing clinical techniques but offers advantages in speed and portability.
- Heterogeneous tissue composition significantly affects classification performance.

## Abstract

Re‐excision rates remain high for early‐stage breast cancer patients due to challenges in margin delineation during surgery, such as poorly defined tumor boundaries. Our group has developed a time‐resolved fluorescence and diffuse reflectance (TRF‐DR) spectroscopy system to assess tumor boundaries where tissue composition is heterogeneous. This study assesses system feasibility for classifying tumor in heterogeneous regions using weighted logistic regression and principal component analysis (PCA). A total of 4818 measurements from 73 frozen patient ex vivo samples were evaluated. Tissue measurement areas (1 mm × 1 mm) were pathologist‐assigned percentages of low (< 25%), medium (25%–< 75%), and high (≥ 75%) for tumor, fibroglandular, or adipose composition. The model trained with four principal components (PCs) achieved sensitivity of 77% and specificity of 68%, comparable to current clinical techniques while being fast, portable, and economical. The TRF‐DR system with weighted logistic regression demonstrates potential as an intraoperative margin assessment tool.

This study demonstrates our time‐resolved fluorescence and diffuse reflectance spectroscopy system can classify breast tumor in heterogeneous tissue. It achieved sensitivity and specificity comparable to other techniques and highlights the effect of heterogeneous measurements on performance. These results support the potential of this system as an intraoperative margin assessment tool.