Abstract
Photoacoustic (PA) imaging is a promising modality for medical diagnostics and therapeutic monitoring, but accurate quantification of contrast agents (CAs) remains a challenge due to nonlinear signal responses and spectral shifts at varying concentrations. These limitations hinder its clinical utility in applications such as tumor detection and treatment monitoring. This study introduces a spectral decomposition method to improve absolute CA concentration estimation in PA imaging. By using a reference spectral library, the approach corrects for signal distortions and nonlinear behavior, overcoming key limitations of traditional intensity-based methods. Validation was performed through in vitro experiments using a prostate-specific membrane antigen (PSMA)-targeted CA in both saline and blood, as well as dynamic tracking of indocyanine green (ICG) in ex vivo tissue. The method achieved significantly lower concentration estimation errors, with average absolute errors of 1.80 µM in saline and 3.34 µM in blood. Compared to conventional techniques, the proposed method demonstrated enhanced reliability and robustness. These results underscore the potential of this spectral-based quantification technique to support more precise, clinically translatable PA imaging, enabling accurate CA measurement for early disease detection, surgical guidance, and real-time monitoring of therapeutic interventions.