Transfer learning as an AI-based solution to address limited datasets in space medicine. (February 2023)

• Record Type: Journal Article
• Authors: Waisberg, Ethan; Ong, Joshua; Kamran, Sharif Amit; Paladugu, Phani; Zaman, Nasif; Lee, Andrew G.; Tavakkoli, Alireza
• Is Part Of: Life sciences in space research. Volume 36(2023)
• Page Start: 36
• Page End: 38
• Publication Date: 2023-02
• Subjects: Transfer learning -- Space medicine -- Machine learning -- Small datasets; Space biology -- Periodicals; 571.0919
• DOI: 10.1016/j.lssr.2022.12.002
• Languages: English
• ISSNs: 2214-5524

Abstract

Highlights: Artificial intelligence (AI) has made a revolutionary impact in many aspects in medicine and is beginning to be integrated into clinical practice. Limited astronaut medical dataset is currently one of the largest barriers for training machine learning algorithms. Transfer learning is a potential solution to overcome this dataset size limitation and can help improve training time and performance of a neural networks. Abstract: The advent of artificial intelligence (AI) has a promising role in the future long-duration spaceflight missions. Traditional AI algorithms rely on training and testing data from the same domain. However, astronaut medical data is naturally limited to a small sample size and often difficult to collect, leading to extremely limited datasets. This significantly limits the ability of traditional machine learning methodologies. Transfer learning is a potential solution to overcome this dataset size limitation and can help improve training time and performance of a neural networks. We discuss the unique challenges of space medicine in producing datasets and transfer learning as an emerging technique to address these issues.