MODELLING AND VERIFICATION OF THERMOACOUSTIC MEDICAL IMAGING FROM NANOSCOPIC TO MACROSCOPIC RESOLUTIONS

Abstract

In this thesis, three main questions regarding the potential of thermoacoustic imaging are answered: 1) what are the conventional resolution limitations of photoacoustic imaging and how can they be extended to enable high-resolution imaging, 2) Can photoacoustic imaging resolution be brought down to nanoscopic levels, and 3) As laser based photoacoustic imaging has been deployed with great success, is it also possible for other radiation to generate useful ultrasound signals for imaging? Whereas laser-induced photoacoustic tomography has been widely explored for a diverse range of biomedical contexts, there remain some fundamental limits to the resolution levels in which it can operate. Namely, the axial resolution of photoacoustic imaging remains restricted by the fact that ultrasonic transducers are not able to detect high-frequency signals that encode nanoscale resolution information. Therefore, there is a lingering question about how photoacoustic imaging can truly enter the realm of nanoscale imaging, as has been done by other modalities such as STED microscopy, structured illumination microscopy, and STORM microscopy. It is believed that laser-based detection in lieu of a transducer may enable a super-resolution photoacoustic imaging modality. However, there remain important questions about the reach and feasibility of nanoscale photoacoustic imaging. Specifically: will highly focused lasers directed at single cells result in thermal damage of biological samples? Will the axial imaging resolution of laser based detection truly be able to overcome the conventional optical diffraction limit of ~200nm? Will optical detection be sensitive enough to detect photoacoustic signals? Consequently, models are developed for thermoacoustic imaging for nanoscale imaging at super-resolutions exceeding that of the optical diffraction limit (~200nm), that show the potential for thermoacoustic imaging to enable super-resolution imaging of single cells. The models confirm that such imaging is possible while simultaneously ensuring the thermal safety of cells as the laser-induced temperature rise of such imaging is only within mK, potentially allowing for high-resolution imaging in vivo. It is also confirmed that a laser of 7ps duration should generate frequencies high enough to enable super-resolutions. Models are also developed for the estimation of the sensitivity and resolution of these high-resolution imaging, and it is predicted that super-resolution photoacoustic imaging may be able to image at axial resolutions of 10nm at noise equivalent number of molecules of 292 in the case of imaging hemoglobin in red blood cells. A length-scale and time-scale generalizable simulation workflow is developed and deployed to generate simulated images of super-resolution photoacoustic imaging, showing the potential of 3D super-resolution achievable via thermoacoustic imaging. This numerical simulation workflow is generalizable to multiple length scales as well as to other sources of radiation. The model predictions regarding detectable high frequency photoacoustic signal generation is experimentally confirmed via the creation and testing of a pump-probe based preliminary photoacoustic imaging system. The system is shown to be capable of detecting a clear and repeatable signal. Acquired A-lines from this system confirm that GHz frequencies can be detected using pump-probe detection in photoacoustics, thereby opening the door for nanoscale photoacoustic imaging However, the experimental results also demonstrate that feasible and convenient nanoscale imaging will require a more stable laser than is available, as pulse to pulse intensity fluctuations in the laser greatly limit the imaging speed and necessary number of averages for a single A-line scan. The developed models show promise and use towards the development of novel thermoacoustic imaging modalities and can be deployed to assess feasibility of different configurations of thermoacoustic imaging prior to the expenditure of resources on experimental realization. In this way, the developed models have the potential to enable the development of various thermoacoustic imaging modalities via a single generalizable framework through which imaging characteristics can be predicted at multiple length and time scales.