The rapid advancement of artificial intelligence (AI) in the healthcare sector, particularly in radiology, has led to the widespread adoption of AI-based diagnostic systems in hospitals. However, concerns have been raised about the environmental impact of complex AI models.
To address this issue, Associate Professor Daiju Ueda of Osaka Metropolitan University‘s Graduate School of Medicine, along with other experts, conducted a research review on the environmental costs of AI in the medical field. The review focused on energy consumption, carbon emissions of data centers, and electronic waste issues.
The team proposed potential solutions, such as developing energy-efficient AI models, implementing green computing, and utilizing renewable energy to mitigate these environmental impacts.
Furthermore, the review suggests steps for the responsible implementation of AI in healthcare. These guidelines are crucial for medical facilities, decision-makers, and AI professionals to utilize AI technology in an environmentally conscious way.
“AI has the potential to improve the quality of healthcare, but at the same time its environmental impact cannot be ignored. The best practices we have recommended are the first steps toward balancing these two factors,” stated Professor Ueda.
“The challenge for the future will be to verify and further elaborate these recommendations in actual medical practice. They are also expected to contribute to the standardization of methods for assessing AI’s environmental impact and the development of an international regulatory framework.”
Journal reference:
- Daiju Ueda, Shannon L Walston, Shohei Fujita, Yasutaka Fushimi, Takahiro Tsuboyama, Koji Kamagata, Akira Yamada, Masahiro Yanagawa, Rintaro Ito, Noriyuki Fujima, Mariko Kawamura, Takeshi Nakaura, Yusuke Matsui, Fuminari Tatsugami, Tomoyuki Fujioka, Taiki Nozaki, Kenji Hirata, Shinji Naganawa. Climate change and artificial intelligence in healthcare: Review and recommendations towards a sustainable future. Diagnostic and Interventional Imaging, 2024; DOI: 10.1016/j.diii.2024.06.002