SBIR Phase I: Artificial Intelligence (AI) Enabled 3-Dimensional (3D) Imaging for COVID-19

这个小企业创新研究第一阶段项目的更广泛的影响/商业潜力是通过提供成像信息来评估疾病的严重程度和进展，从而改进新冠肺炎的治疗。通过对磨玻璃样阴影的观察，胸部计算机断层扫描(CT)已被证明对新冠肺炎敏感，目前正用于有急性症状的患者。在决定使用CT成像时，必须考虑实际的考虑因素，如对患者的高辐射和移动患者进行成像所带来的交叉污染风险。低成本、低辐射和移动性的3D成像解决方案可以提供优势并带来高质量的护理，同时有效地集成到医院工作流程中。除了应对当前的危机，使用该解决方案来应对其他呼吸系统疾病将为这项研究确保强大的商业潜力。这个小企业创新研究(SBIR)第一阶段项目旨在开发和验证一种启用人工智能(AI)的3D成像重建算法，该算法可用于评估新冠肺炎等呼吸系统疾病的严重程度和进展。目前的胸部成像技术可以提供足够的图像质量或有效的肺部成像，但不能两者兼而有之。两个主要的进步可以使成像更有效率，同时也提供所需的图像质量。散射建模已被证明在从几张射线照片重建时成功地改善了图像质量；初步数据显示了如何结合机器学习(ML)来增强高效成像以提供更高质量的图像。一种模拟X射线散射并使用ML从快速射线照片重建3D图像的3D图像创建算法将使3D成像能够用于包括新冠肺炎在内的呼吸系统疾病。该算法将在身体模型上进行验证，以评估移动的、可在床边使用的、易于覆盖以进行无菌使用的人工智能成像系统是否可行。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 Innovation Research (SBIR) Phase I project is to improve COVID-19 treatments by providing imaging information to assess the severity and progression of the disease. Chest computed tomography (CT) has been shown to be sensitive to COVID-19 via the observation of ground-glass opacities and is being used on patients with acute symptoms. Practical considerations such as high radiation to the patient and cross contamination risk from moving the patient for imaging must be taken into account in the decision to image with CT. A 3D imaging solution that is low cost, low radiation, and mobile could provide advantages and bring quality of care while integrating efficiently in the hospital workflow. Along with addressing the current crisis, the usage of this solution to address other respiratory diseases would secure strong commercial potential for this research.This Small Business Innovation Research (SBIR) Phase I project seeks to develop and validate an artificial intelligence (AI)-enabled 3D imaging reconstruction algorithm that can be used to assess the severity and progression of respiratory diseases such as COVID-19. Current chest imaging technologies can either provide adequate image quality or efficient imaging of the lungs, but not both. Two major advances could make the imaging more efficient while also providing the required image quality. Scatter modeling has been shown to be successful in improving image quality when reconstructing from few radiographs; Preliminary data shows how Machine Learning (ML) can be integrated to enhance efficient imaging to provide higher quality images. A 3D image creation algorithm that models X-ray scatter and uses ML to reconstruct 3D images from rapid radiographs will enable using 3D imaging for respiratory diseases including COVID-19. This algorithm will be validated on cadaveric models to assess if an AI-enabled imaging system that is mobile, that can be used bedside, and that is easily draped for sterile utilization is feasible.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.