A recent study in South Korea explored the use of a cryptographic scheme widely used in secure elections to protect patient data used in training AI models.
Researchers from Asan Medical Center applied homomorphic encryption (HE) to big data collected from many organizations to test predictive AI models. HE is an encryption method that allows computation on encrypted data.
RESULT
In one study, there were The findings, published in JMIR Medical Informatics, involved collecting EMR data from more than 300,000 patients from three hospitals – Asan Medical Center, Seoul National University Hospital, and Ewha Womans University Seoul Hospital – to train an AI model to predict 30-day mortality after surgery. The data was encoded in HE.
Research has demonstrated that HE can be practically applied to combine large volumes of data from multiple sources while ensuring privacy and security.
The findings also suggest that small hospitals can leverage this coding approach to develop their own AI models by securely mining data from larger hospitals. These models have better predictive capabilities than models developed based on raw data from a single institution.
WHY IT MATTERS
More than a decade since HE was demonstrated in a thesis in the United States, many highly regulated fields such as finance, IT, and healthcare have explored its application to enhance data privacy. Even in elections, this encryption method is increasingly applied to ensure accurate poll results.
Research from South Korea also looks at HE, attempting to overcome current limitations in healthcare AI development. While having larger, more diverse data is key to improving the accuracy and applicability of AI models, collecting it is challenging due to strict privacy regulations.
“In this multicenter study, we used advanced HE to protect against personal information leakage and data security. In addition, HE enables predictive operations and modeling on encrypted data, providing an optimal solution that can completely solve the problems related to personal information leakage and data security. Furthermore, HE provides ‘strongest’ security when used appropriately, such as in outsourced computing, where HE protects against data breaches in the computation,” the researchers noted in their findings.
THE BIGGER TREND
The South Korean government has supported local medical institutions in pursuing AI research by facilitating collaboration. For example, the Ministry of Health and Welfare launched the Medical Data Utilization Project, which aims to help digital health researchers connect with hospitals and access their data sets. Last year, the ministry designated five Korean hospitals as centers for safe medical data utilization.
