Application of multimodal optical imaging for brain function mapping and liver fibrosis monitoring

Abstract

Optical imaging has increasingly gained prominence as a transformative modality in biomedical research due to its unique capacity for non-invasive imaging with high resolution and molecular specificity. Traditional medical imaging techniques, such as X-ray imaging, computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), and ultrasound (US), are widely utilized but are inherently limited in resolving cellular-level structural and functional details. Pure optical imaging methods, despite achieving high resolution, are significantly affected by strong tissue scattering, which greatly limits imaging depth and induces an urgent need for techniques that overcome these barriers. Emerging optical imaging methods, notably photoacoustic imaging (PAI), have been developed to overcome these limitations. PAI relies on optical absorption to provide intrinsic optical contrast of biological tissues, significantly overcoming the influence of optical scattering. This hybrid technology offers deeper penetration and superior contrast by combining optical excitation with acoustic detection. PAI leverages the photoacoustic effect, in which absorbed pulsed laser energy in biological tissues induces thermoelastic expansion, generating acoustic signals detectable by ultrasound transducers. This hybrid modality surpasses traditional optical imaging methods by providing enhanced imaging depth, reduced scattering effects, and greater sensitivity to endogenous and exogenous contrast agents. Multiscale photoacoustic technologies, including photoacoustic computed tomography (PACT), photoacoustic microscopy (PAM), and photoacoustic endoscopy (PAE), have demonstrated promising applications in neuroscience, oncology, and hepatology. Recently, PAI has demonstrated clear advantages and extensive research potential in areas such as brain function mapping and liver disease detection. In brain function mapping, it provides valuable parameters including cerebral electrophysiological activities, oxygenation levels, and hemodynamic responses, which is essential for neurological research and disease diagnostics. Regarding liver disease detection, PAI effectively assesses microvascular structures, tissue oxygenation, and biomechanical properties, significantly contributing to early diagnosis and monitoring of liver conditions such as fibrosis and hepatocellular carcinoma. This thesis specifically explores two significant biomedical applications of PAI: brain function mapping and liver fibrosis detection. In the first part, brain function mapping, we demonstrate PAI's capability to perform label-free, high-resolution imaging of cerebral electrophysiological activities, essential for diagnosing neurological disorders such as epilepsy. Our work is divided into three parts. Firstly, cellular experiments validated the optical response mechanism of voltage-sensitive dyes (VSD), demonstrating opposite changes in photoacoustic and fluorescence intensities in relation to membrane potentials, adhering strictly to energy conservation. Secondly, we utilized in-vivo fluorescence imaging to track VSD signals in epileptic mouse models, aligning with our cellular observations. However, the relatively low temporal and spatial resolution limited the signal mapping to minute-level temporal dynamics across the whole brain, restricting precise microscopic localization and detailed observation of epilepsy. Thirdly, PACT was employed, achieving a high-resolution (approximately 100 µm) whole-brain coverage (5.5 cm width) with substantial imaging depth (down to 4 mm) and a high imaging frame rate (up to 10 Hz). Our findings revealed that this method directly tracked epileptic signals, accurately localized epileptic foci within specific brain regions, and demonstrated the transmission and interaction among these epileptic foci. Through this innovative VSD technique combined with PAI (PA-VSD), real-time monitoring and precise localization of epileptiform discharges were achieved, providing a potential tool for intraoperative guidance and therapeutic intervention. The second part focuses on liver fibrosis detection, presenting a novel photoacoustic elastomicroscopy (PAEM) system that integrates structural vascular imaging and biomechanical assessment via time-of-flight (ToF) quantification — a method particularly well-suited to the liver due to its rich vascularization and hemoglobin content, which enhance compatibility with photoacoustic monitoring. The exploration was divided into two parts. Firstly, we utilized high-resolution PAM to image liver tissue sections and found that significant alterations in vascular architecture—especially hepatic lobule structures—were already evident at the early stages of fibrosis. This finding aligns with existing literature, which indicates that structural changes precede biomechanical property alterations. However, we observed that structural imaging alone could only distinguish early fibrosis (0-3 weeks), but it failed to effectively differentiate middle to late stages. Therefore, we proposed a second component utilizing ToF-based elasticity quantification derived from ultrasound propagation speed through tissue to infer relative stiffness. Although this does not yield absolute mechanical values, it enables quantitative comparisons across the full spectrum of fibrosis (0-16 weeks). This approach offers clear advantages in early-stage detection and addresses the limitations of shear wave elastography (SWE), which lacks sufficient sensitivity for early-stage diagnosis. This methodology allows early, sensitive detection of subtle hepatic changes, surpassing conventional imaging techniques in accurately staging fibrosis progression. By coupling structural vascular analysis with biomechanical quantification, this approach offers a comprehensive diagnostic platform for chronic liver disease management as well as the potential in hepatic tumor margins. The presented work underscores the translational potential of photoacoustic modalities in clinical diagnostics and personalized medicine, establishing a methodological foundation for future advancements in biomedical imaging. Specifically, in the context of brain function mapping, the integration of PA-VSD paves the way for non-invasive, real-time neurophysiological monitoring with both spatial specificity and temporal precision—capabilities highly desirable for epilepsy surgery, brain-computer interface development, and broader neurological disease management. Similarly, in hepatology, the development of PAEM for concurrent structural and biomechanical assessment presents a promising solution for the early detection and quantitative staging of liver fibrosis. As liver fibrosis remains clinically underdiagnosed in its initial stages, the sensitivity and depth-resolved precision offered by PAI hold great potential for improving patient outcomes. With further optimization and validation, the methods presented in this thesis may accelerate the translation of PAI into routine clinical workflows, offering clinicians robust tools for real-time, label-free, and multi-parametric disease evaluation.