Medical technology company Medical IP has developed an AI-driven whole-body CT technology that can predict the incidence and prevalence of type 2 diabetes while identifying other cardiometabolic issues using CT scans.
Founded in 2015 by Joon S. Park, the company has developed DeepCatch, an AI-powered software that leverages deep learning algorithms to automatically analyze body composition data from CT scan images to assess an individual’s risks of diabetes and related conditions for early medical intervention.
According to a study conducted in partnership with Kangbuk Samsung Hospital and Seoul National University Hospital, the software can be used to comprehensively assess the risks of diabetes and related cardiometabolic conditions.
A Promising Result
According to a report published by Korea Biomedical Review, the software analyzed scans from 32,166 participants aged 27 to 83 and found that the visceral fat index was a major indicator of diabetes incidence and prevalence.
The AI software also showed a significantly high accuracy in diagnosing related conditions such as sarcopenia, coronary artery calcification, osteoporosis, and fatty liver.
The study finding shows that CT scans can be vital in preventive screening for other medical conditions associated with diabetes, extending its use beyond the traditional disease diagnosis, said Professor Chang Yoo-soo of Kangbuk Samsung Hospital.
With the new AI-powered diagnostic software, the researchers are confident that CT images could soon be used to screen for diabetes and other cardiometabolic conditions simultaneously, allowing medical professionals to predict diabetes onset in healthy individuals. This can also allow for early detection and prevention of complications.
Additionally, the new technology will allow individuals to obtain biomarker information for metabolic conditions and other body composition-related illnesses during their medical check-ups from just a single CT scan.
