IRES Track I: International Research Experience for Students on Artificial Intelligence for Congenital Heart Diseases

该IRES项目的目的是为美国学生提供宝贵的研究经验，在位于广东省人民医院（GGH）的广东省心血管病研究所（GCI）探索用于先天性心脏病（CHDs）的新型人工智能（AI）算法。位于中国广州的广州总医院是中国最大的医院之一，其附属的GCI在心血管疾病研究方面一直排名前三。该项目每年将在全国范围内选择三（3）名研究生和三（3）名本科生，并支持他们在十二（12）周内访问GCI，与世界知名的心血管医学研究人员在身临其境的环境中合作，通过创新的人工智能算法解决一些最关键和最具挑战性的问题，这些算法可以使大量患有CHD的人群受益。CHD是出生缺陷是导致死亡的主要原因，在美国每110个新生儿中就有1个出生缺陷，在发展中国家甚至更多。CHD具有非常大的结构变化，这使得手动诊断耗时且繁琐。此外，冠心病最常见的治疗方法是心脏手术，这是相当复杂的，被广泛认为是外科皇冠上的宝石。目前，只有放射科医生和外科医生在冠心病方面有丰富的经验，才能自信地处理它们。该项目的结果将潜在地提高冠心病治疗的有效性和效率，并使全球大量患有冠心病的人群受益。该项目还将使美国学生能够与他们在GCI的研究导师合作，对CHD的AI进行高质量的研究。这些经历使他们在职业生涯的关键早期阶段接触到国际研究界。通过参与该项目，美国学生将获得关于深度学习与医疗应用研究、国际医疗保健前景、中国文化以及在国际环境中执行和合作的广泛经验。这些经验还将通过个人社交媒体和外展活动分享给更广泛的社区，例如为对AI感兴趣的全国K-12学生举办的KidAI网络研讨会。为确保可持续发展，该计划将为参与学生提供专业发展支持。IRES项目的科学目标是探索创新的人工智能算法，有助于冠心病的诊断、治疗计划和预后。该研究通过GCI的医生和IRES访问工程领域的学生之间的合作，针对医学图像处理和冠心病诊断中迫切需要解决的基本问题。虽然人工智能在普通医学中显示出巨大的潜力，但CHD中的大心脏结构和大血管连接变化无法通过为其他疾病开发的现有人工智能算法来处理。拟议的研究将是第一个应对这一挑战的研究，预计将开发创新算法。可以预见的是，通过与中国心血管疾病领域领先研究团队的密切互动，GCI的这些项目所取得的突破将对基于人工智能的医学研究、临床应用和社会公益产生直接影响。该奖项反映了NSF的法定使命，并通过基金会的智力价值和更广泛的影响评审标准进行评估，被认为值得支持。

The objective of this IRES project is to provide U.S. students with valuable research experience in the exploration of novel artificial intelligence (AI) algorithms for congenital heart diseases (CHDs) in Guangdong Provincial Cardiovascular Institute (GCI), located at Guangdong General Hospital (GGH). Located in Guangzhou, China, GGH is one of the biggest hospitals in China, and the GCI affiliated with it is constantly ranked top three in terms of cardiovascular disease research. The project will select three (3) graduate students and three (3) undergraduate students nation-wide each year and support them to visit GCI over a period of twelve (12) weeks to work with world-renowned researchers on cardiovascular medicine in an immersive setting, addressing some of the most critical and challenging problems through innovative AI algorithms that can benefit the large population suffering from CHD.CHD is the leading cause of mortality from birth defects, which occurs 1 in every 110 births in the United States, and even more in developing countries. CHD comes with very large structure variations which makes it time-consuming and tedious to diagnose manually. In addition, the most common treatment for CHD is cardiac surgery, which is rather complex and is widely regarded as the jewel in the crown of surgery. Currently, only radiologists and surgeons with extensive experience in CHD can handle them confidently. The outcome of this project will potentially boost the effectiveness and efficiency of CHD treatments, and benefit the large population worldwide who suffer from CHD. The project will also enable U.S. students to conduct high-quality research on AI for CHD, in collaboration with their research mentors at GCI. Such experiences expose them to the international research community at a critical early stage in their careers. It is expected that through participating in this program, U.S. students will gain extensive experience on the research of deep learning with medical applications, on the international healthcare prospects, on Chinese culture, and on performing and collaborating in an international environment in general. The experience will also be shared to the broader community through personal social media as well as outreach events such as KidAI webinars for national K-12 students interested in AI. To ensure sustainability, the project will provide professional development support for the participating students. It will also focus on minorities and underrepresented groups.The scientific objective of this IRES project is to explore innovative AI algorithms that can benefit the diagnosis, treatment planning and prognosis of CHDs. The proposed research, through the collaborationbetween doctors at GCI and the IRES visiting students in the engineering fields, targets fundamentalissues in medical image processing and diagnosis of CHD that need to be addressed urgently. Although AI has shown great potential in general medicine, the large heart structure and great vessel connection variations in CHDs cannot be handled well by existing AI algorithms developed for other diseases. The proposed research will be the first to tackle this challenge and innovative algorithms are expected to be developed. It can also be anticipated that with the close interaction with leading research group on cardiovascular diseases in China, the breakthroughs made from these projects at GCI can have a direct impact on AI-based medicine research, clinic usage, and social welfare.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.