I-POCUS: An Intelligent Point-of-Care Ultrasound System for Neonatal Intensive Care Units in Canada's Hospitals

医疗点超声(PICUS)系统可帮助医生在任何情况下对患者进行快速、准确和安全的成像。超声(US)扫描是新生儿重症监护病房(NICU)使用的主要成像方式，因为任何其他侵入性成像程序在新生儿中都不可行。虽然Pocus的使用已被认为能够提供准确的信息，以帮助临床医生诊断新生儿的心肺疾病，但目前这项技术主要由少数临床医生在大型学术中心应用，原因是相对广泛的培训要求和获得大量患者人口以保持技能的障碍。从美国图像中评估和跟踪心脏和肺功能障碍需要大量的专业知识和培训，以便获得足够高质量的图像。引导学员获取高质量图像的智能辅助系统是其中的关键组成部分。随着时间的推移，还需要开发智能算法来自动化心肺功能障碍的评估、诊断和跟踪。 基于我们最近与该提案的工业合作伙伴Dapasoft Inc.的成功合作，这个由工程、计算机科学和医学领域的成员组成的多学科团队提议研究开发智能Pocus(I-Pocus)系统的方法和技术，用于培训在NICU操作的医疗专业人员。具体地说，基于人工智能(AI)最近的成功，该项目的目标是在图像和视频处理、数据分析和可视化方面开发新颖和智能的方法和算法，以培训能够在NICU使用I-Pocus系统的医疗专业人员，以便智能和可靠地诊断新生儿心肺疾病，并为关键的临床问题提供解决方案。

Point-of-Care Ultrasound (POCUS) systems help medical practitioners in obtaining fast, accurate, and safe imaging of patients in any scenarios. Ultrasound (US) scan is the principal modality of imaging being used in Neonatal Intensive Care Units (NICUs), since any other invasive imaging procedure is not feasible in neonates. Although the use of POCUS has been recognized to be able to provide accurate information to assist clinicians in diagnosing neonatal heart and lung conditions, currently this technology is mostly applied in large academic centres by a handful of clinicians, due to the barriers of relatively extensive training requirement, and access to a large patient population to maintain skills. Assessment and tracking of cardiac and lung dysfunctions from US images require significant expertise and training so that images of sufficient quality can be obtained. An intelligent assistant system to guide trainees in acquiring quality images is a key component. The development of intelligent algorithms is also required to automate the assessment, diagnosis, and tracking of cardiac and lung dysfunctions over time. Based on our recent successful collaboration with Dapasoft Inc, industrial partner of the proposal, this multidisciplinary team with members in engineering, computer science and medicine propose to investigate methods and techniques for the development of an intelligent POCUS (I-POCUS) system for the training of medical professionals operating in NICUs. Specifically, riding on recent success of artificial intelligence (AI), the target of the project is to develop novel and intelligent methods and algorithms in image and video processing, data analytics and visualization for the training of medical professionals capable of using of the I-POCUS systems in NICUs in order to intelligently and reliably diagnose neonatal heart and lung conditions and provide solutions to key clinical problems.