Ultra-high spatial resolution photon-counting CT with multiple focal spots

PROJECT SUMMARY / ABSTRACT The emergence of clinical photon-counting CT shows much promise for new clinical diagnostics owing to its noise advantages, improved material discrimination, and small detector pixels. The latter element has the po- tential to have particular impact on the overall spatial resolution of the system. Recently available high-resolution systems have demonstrated particular advantage in cardiothoracic imaging and pulmonary imaging in particular, where good visualization of the fine structure of the lung can have significant impact on diagnostics. These x-ray detector developments have, in some ways, outpaced x-ray source developments; and, with these new systems the x-ray now tends to be the limiting factor for high-resolution capability. The fundamental issue is that smaller detector voxels require a smaller x-ray focal spot – which tends to limit fluence and increase overall noise. This compounds with the general need for more fluence when image voxels are made smaller (higher resolution) to keep from increasing noise. In x-ray tube design there is an inherent trade-off where smaller focal spots have lower capacity for generating fluence. Thus, much of the capability of high-resolution CT with photon-counting detectors is unrealized. We proposed a novel data acquisition strategy that uses multiple focal spots. That is multiple, structured spots that change the balance of focal spot size and fluence – e.g. performing a data acqui- sition with both a large and small focal spot to get both high-resolution information but sufficient fluence to reduce noise. Combined with state-of-the-art model-based and deep learning approaches, we will develop a new para- digm for ultra-high-spatial resolution CT (UHR-CT). We seek to accomplish that development through the follow- ing specific aims: Aim 1: Develop system models and reconstruction approaches for data acquisition with multiple focal spots, in which high-fidelity model form a framework for both simulation and joint data processing of the multiresolution data associated with the multiple focal spot technique. Aim 2: Investigate and evaluate different multiple focal spot strategies for UHR-CT, wherein we study a range of system designs from current technology to more complex focal spot designs, and evaluate optimized strategies in simulated and physical systems. Aim 3: Assessment of UHR-CT in clinical PCCT on lifelike phantoms and an in-vivo animal model, where we translate the multiple focal spot method to a clinical PCCT for immediate impact. Successful completion of these aims will demonstrate the underlying technology, validate the methods, and characterize the potential for high-resolution performance improvements using both current and emerging x-ray source technol- ogies. The availability of the ultra-high-resolution capability opens the doors to a wide range of potential clinical diagnostic as smaller and smaller features can be resolved.