Abstract

Cancer remains a leading cause of death worldwide, underscoring the need for imaging technologies that provide both structural and functional insights into cancer biology. Photoacoustic imaging (PAI), which detects ultrasonic signals generated by optical absorption, combines the strengths of both modalities, achieving optical contrast with ultrasonic resolution and depth. When co-registered with ultrasound (US), the hybrid US-PAI platform offers a unique combination of high-resolution anatomical and functional imaging that can improve cancer diagnosis, characterization, and treatment monitoring. This dissertation explores the clinical translation of US-PAI through innovations in imaging system design, computational analysis, and clinical validation across three major cancer types: ovarian, rectal, and breast cancer. In ovarian cancer, quantitative vascular biomarkers from photoacoustic tomography combined with multiparametric and radiomic analysis on co-registered ultrasound enabled improved differentiation between benign and malignant ovarian lesions. In rectal cancer, deep learning analysis of co-registered ultrasound-photoacoustic microscopy provided early indicators of response to neoadjuvant therapy by quantifying vascular remodeling that occurred prior to anatomical recovery. In breast cancer, diffuse optical tomography, photoacoustic tomography, and ultrasound were integrated to jointly reconstruct optical absorption and scattering maps, yielding more accurate quantitation of hemoglobin and oxygenation biomarkers while preserving vascular detail. In addition, this work extended to optical coherence tomography and photoacoustic microscopy for high-resolution microvascular and morphological imaging of endometrial cancer, exploring both conventional computer vision methods and deep learning based analysis to characterize endometrial tissue microarchitecture. Together, the studies presented in this dissertation establish a consistent imaging framework that correlates PAI-derived functional parameters with underlying cancer biology across different organ systems and cancer types. By combining structural and functional imaging within a single clinically translatable platform, the work demonstrates the potential of US-PAI to improve diagnostic accuracy, enable early treatment assessment, and support precision oncology.

Committee Chair

Quing Zhu

Committee Members

Chao Zhou; Ian Hagemann; Matthew Lew; Song Hu

Degree

Doctor of Philosophy (PhD)

Author's Department

Biomedical Engineering

Author's School

McKelvey School of Engineering

Document Type

Dissertation

Date of Award

12-9-2025

Language

English (en)

Download

Share

COinS