Accurate identification and mechanism of breast invasive ductal carcinoma based on combining steady-state and time-resolved label-free fluorescence spectroscopy.

Breast cancer (BC) remains the most prevalent malignant tumor and a major threat to women's health. Frozen section examination, a standard method for margin assessment during breast-conserving surgery (BCS), is time-consuming and highly dependent on pathologist expertise. This study investigates the feasibility of using laser-induced fluorescence (LIF) spectroscopy to classify ex vivo breast tissues. Tissues from 17 patients were examined using steady-state and time-resolved fluorescence spectroscopy. A new spectral fluorescence intensity ratio (FIR) was calculated and analyzed combining with the mean fluorescence lifetime (MFL). Linear discriminant analysis (LDA) was applied to assess classification accuracy. The results indicate that the intrinsic fluorescence of breast tissue can be attributed to NAD(P)H, FAD, collagen, and elastin. Carcinoma tissues showed statistically lower FIR and shorter MFL (P < 0.001) than normal tissues. Changes in steady-state spectra might reflect elevated NAD(P)H levels and variations in binding ratios. Meanwhile, the observed shortening of fluorescence lifetime could potentially result from a reduced fluorescence contribution of structural proteins and a decreased proportion of bound NAD(P)H. Combining the fluorescence intensity ratio and MFL within an LDA framework significantly improved diagnostic accuracy and enabled reliable classification of ex vivo breast carcinoma and normal tissues. This spectroscopic approach offers a label-free, quantitative tool with potential for real-time margin assessment during BCS, aiding precise intraoperative decision-making.