STTR Phase I: AI-assisted Assessment, Tracking, and Reporting of COVID-19 Severity on Chest CT

这个小企业技术转让第一阶段项目的更广泛的影响/商业潜力是利用人工智能(AI)来减少错误，并提高在胸部CT图像上评估新冠肺炎肺部疾病严重程度的准确性、标准化、一致性和报告。胸部CT检查在新冠肺炎患者中起着至关重要的作用，但目前评估胸部CT图像的方法缺乏准确、定量或一致的信息，导致基于文本的报告难以解读。拟议的人工智能辅助新冠肺炎胸部CT工作流程将通过提供标准化的图形报告、重要发现的关键图像和结构化文本，有效地捕获肺部受累的部分，并改善与临床医生的沟通。这些量化数据将标准化个别患者的报告，并为人群水平的分析提供数据，从而为在国家层面大幅推进对新冠肺炎肺部疾病的科学知识提供潜力。该项目第一阶段建议开发一个人工智能辅助的新冠肺炎胸部CT工作流程，以快速和客观地量化肺部受累的百分比，使用新冠肺炎报告和数据系统(CO-RADS)对肺部受累进行分类，跟踪常见和不常见的新冠肺炎肺部发现，并自动生成包含图表、关键图像和结构化文本的总结报告。在胸部CT图像上评估和报告新冠肺炎肺部疾病严重程度的护理标准涉及基于文本的口述报告，这些报告主观、高度可变、生成和解释效率低、容易出错、不完整和定性，使用非标准化格式提供的数据。拟议的人工智能辅助新冠肺炎胸部CT工作流程将减少口译错误和遗漏，并提高新冠肺炎疾病严重程度评估和治疗反应的准确性、标准化、观察员之间的一致性、效率和报告。该项目将与专家临床医生团队一起验证工作原型。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact /commercial potential of this Small Business Technology Transfer (STTR) Phase I project is to leverage artificial intelligence (AI) to reduce errors and improve accuracy, standardization, agreement, and reporting in evaluation of COVID-19 lung disease severity on chest computed tomography (CT) images. Chest CT procedures play a critical role in COVID-19 patients but current methods for evaluating chest CT images lack accurate, quantitative, or consistent information, leading to text-based reports that are difficult to interpret. The proposed AI-assisted COVID-19 chest CT workflow will efficiently capture the fraction of lung involvement and improve communication with clinicians by providing a standardized graphical report, key images of important findings, and structured text. The quantitative data will standardize reporting on an individual patient basis and provide data for population-level analyses, thereby offering the potential to significantly advance scientific knowledge of COVID-19 lung disease on a national level. This STTR Phase I project proposes to develop an AI-assisted COVID-19 chest CT workflow to rapidly and objectively quantify the percentage of lung involvement, classify lung involvement using the COVID-19 Reporting and Data System (CO-RADS), track common and uncommon COVID-19 lung findings, and automatically generate summary reports with a graph, key images, and structured text. The standard-of-care for assessing and reporting COVID-19 lung disease severity on chest CT images involves dictated text-based reports that are subjective, highly variable, inefficient to generate and interpret, prone to errors, incomplete, and qualitative with data provided in an unstandardized format. The proposed AI-assisted COVID-19 chest CT workflow will reduce interpretation errors and omissions and improve accuracy, standardization, inter-observer agreement, efficiency, and reporting in evaluation of COVID-19 disease severity and response to treatment. This project will validate the working prototype with a team of expert clinicians.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.