Resolution Enhancement in Acoustic-Resolution Photoacoustic Microscopy using Measured Impulse Response Function (MIRF)-based Reconstruction

Abstract

Established clinical imaging modalities such as ultrasound, computed tomography, and magnetic resonance imaging provide excellent anatomical detail but are inherently constrained in visualizing functional or molecular processes in real time without exogenous contrast agents. Photoacoustic imaging (PAI) has emerged as a hybrid modality that bridges this gap by combining optical absorption contrast with the deep acoustic penetration of ultrasound, enabling non-invasive visualization of vascular structures and oxygen saturation through endogenous chromophores such as hemoglobin. Acoustic-resolution photoacoustic microscopy (AR-PAM), a form of PAI optimized for deep-tissue imaging, remains constrained by a shallow depth of field and signal distortion caused by the transducer’s intrinsic impulse response (IRF), resulting in compromised resolution. To address these limitations, this dissertation proposes a physics-based reconstruction technique that directly incorporates an experimentally measured impulse response function (MIRF) into the image reconstruction process. Unlike conventional deconvolution-based methods, the proposed approach applies a frequency-band-separated MIRF as a time-domain weighting function to effectively correct for the transducer’s amplitude and phase distortions. To validate this methodology, an AR-PAM system based on a needle hydrophone (NH) was first implemented, leveraging its wide acceptance angle and ultra-broad bandwidth. Imaging experiments using a tungsten wire, a metal bee phantom, a chick embryo (ex Ovo), and a nude mouse (in vivo) demonstrated that the synthetic aperture focusing technique (SAFT)+MIRF reconstruction achieved uniform resolution across depth and improved the signal-to-noise ratio (SNR) in deeper regions compared to conventional SAFT. In contrast, while the coherence factor (CF) technique enhanced lateral resolution and contrast, it compromised axial resolution and suppressed weaker signals in the presence of dominant high-intensity components, hindering the visualization of lower-intensity structures. Subsequently, the MIRF-based reconstruction was extended to an array-transducer-based AR-PAM system to evaluate its clinical applicability. Although data acquisition was performed via mechanical scanning, applying MIRF to the composite dataset synthesized from multiple array elements improved structural clarity and resolution compared with SAFT within the transducer’s effective focal depth, as verified through skeleton leaf phantom imaging. In conclusion, this dissertation demonstrates that an experimentally measured MIRF provides an effective, physics-based solution to mitigate the axial resolution degradation and depth inhomogeneity inherent to conventional AR-PAM. The proposed approach enhances both image quality and quantitative performance in single-element (NH) and array-based systems, establishing a solid foundation for future development of high-resolution, deep-tissue, and clinically translatable photoacoustic imaging platforms.