Automated medical diagnostic image quality control using AI-based techniques

计算机断层扫描（CT）和磁共振成像（MRI）图像是常规用于诊断、治疗计划和监测疾病进展的关键诊断工具。CT和MRI图像对于评估对肺部、大脑和血管造成的损害尤为重要，特别是我们正在继续发现新冠肺炎感染对身体的全身影响。放射扫描仪上患者诊断的准确性和及时性是患者扫描期间产生的图像质量的直接函数。 质量差的图像可能导致漏诊或误诊。在最好的情况下，由错误校准的扫描仪获取的图像将导致昂贵、耗时、重复的患者扫描，并且如果未被检测到，则可能导致漏诊，这可能严重危及患者的健康，并且在某些情况下导致死亡。 这项研究的总体目标是开发一种新的解决方案，利用基于AI的临床图像分析来执行自动质量控制（QC）。通过与不列颠哥伦比亚省大学（UBC）、高级质量系统（AQS）和内政部卫生局（IHA）的合作，拟议的解决方案将使用基于人工智能（AI）的临床图像分析，这将直接提高加拿大CT和MRI扫描仪用于患者诊断、治疗、研究和开发的能力和可用性。随着加拿大从COVID-19大流行中恢复过来，这项研究将有助于在感染控制措施和工作人员倦怠导致医疗人员减少的时期内解决对快速，准确临床诊断的日益增长的需求。

Computed Tomography (CT) and Magnetic resonance imaging (MRI) images are critical diagnostic tools routinely use for diagnosis, treatment planning, and monitoring disease progression. CT and MRI images are particularly critical for evaluating the damage caused to the lungs, brain, and blood vessels, particularly as we are continuing to discover the systemic effects that Covid -19 infections have on the body. The accuracy and timeliness of patient diagnosis on radiology scanners is a direct function of the quality of the images produced during a patient scan. Poor quality images can result in a missed or incorrect diagnosis. In the best-case scenario, an image acquired by a mis-calibrated scanner will result in a costly, time consuming, repeated patient scan, and if undetected it can result in a missed diagnosis which may seriously risk patients' health and, in some cases, lead to death. The overall goal of the proposed research is to develop a novel solution for performing automated Quality Control (QC), utilizing AI-based analysis of clinical images. Through a partnership with the University of British Columbia (UBC), Advanced Quality Systems (AQS), and Interior Health Authority (IHA), the proposed solution will use artificial intelligence (AI)-based analysis of clinical images that will directly increase the ability and availability of Canadian CT and MRI scanners to be used in patient diagnosis, treatment, research and development. As the Canada recovers from the Covid-19 pandemic, this research will help address the increased need for rapid, accurate clinical diagnosis during a period of reduced healthcare staffing levels imposed by infection control measures and staff burnout.