CityU interdisciplinary research on big data and AI improves the precision of clinical and healthcare practices
The research paper CheXMed: A multimodal learning algorithm for pneumonia detection in the elderly, with PhD student Ren Hao as the first author and Assistant Professor Jing Fengshi at City University of Macau (CityU) as the co-first author, was published in the highly regarded SCIE-Q1 journal Information Sciences. This paper examines the use of big data and AI in healthcare and clinical practices, specifically in the detection of pneumonia in elderly patients.
This research, which incorporates big data and AI into healthcare, aligns with the government’s plan for appropriate economic diversification. It presents the potential for interdisciplinary collaboration to bridge the gap between academic achievements and industry advancements.
First Author Ren Hao
Co-first Author Assistant Professor Jing Fengshi
Elderly people are one of the high-risk populations for diseases such as pneumonia. Traditional artificial intelligence assisted diagnostic tools for pneumonia detection using imaging are mainly designed for the general population, thus overlooking the specific characteristics of imaging data in geriatric patients.
To address this issue, CityU researchers collaborated with scholars from Sun Yat-sen University, The University of Hong Kong, University of Warwick, Guangdong Second Provincial General Hospital, and City University of Hong Kong to develop a multimodal learning algorithm called CheXMed. This algorithm takes into account the unique features of imaging data in elderly patients.
The results showed that: 1. The SOTA model designed for the general population performed poorly in detecting pneumonia in the elderly. 2. CheXMed, which incorporates clinical and imaging data from elderly patients, significantly improved the precision of pneumonia detection.
The research findings present significance in pneumonia detection among elderly and pave the way for the development of smart health and big health industry in Macao. Through the use of AI and deep learning technologies, the team was able to provide more precise and efficient clinical detection and medical services to the elderly population, showcasing the team as well as CityU’s commitment to social responsibility and caring services in the field of medicine and healthcare.
In January of this year, at the annual conference of the Radiation Medicine Branch of Guangdong Medical Association held at the Fifth Affiliated Hospital of Sun Yat-sen University in Zhuhai, Assistant Professor Jing Fengshi was invited to give a keynote presentation on the research findings. The presentation showcased the significance of the research, and highlighted the close exchange and cooperation between the Guangdong-Hong Kong-Macao Greater Bay Area, especially between Zhuhai and Macao, in the field of big health.
Information Sciences is a highly regarded academic journal published by Elsevier that focuses on information science, intelligent system applications in informatics and computer science. It is listed in the Q1 in the Computer Science category. Its impact factor (IF) is 8.1 according to Journal Citation Reports (JCR).
Detail: https://www.sciencedirect.com/science/article/pii/S0020025523014391
Source: Research Management Office
