Development status

Software as a Medical Device (SaMD) Utilizing Artificial Intelligence for Supporting Diabetes Treatment


Insulin injection therapy is necessary to strictly control blood glucose levels in diabetes and prevent diabetic complications. However, the safe dose range of insulin is narrow, and the optimal type and dosage must be carefully selected for each patient, since overdosing results in hypoglycemia. On the other hand, since diabetologists account for less than 2% of all physicians and are geographically unevenly distributed, diabetes patients currently do not always see diabetologists as their primary care physicians, but rather often see a non-diabetologists. In collaboration with Tohoku University, NEC Corporation and NEC Solution Innovator, Ltd. (NES), we have developed DM-SAiL (Diabetes Skill Acquisition Learning, SAiL), the software as medical device (SaMD) based on the artificial intelligence (AI), that will help non-diabetologists perform diabetologist-level insulin therapy.

There are approximately 10 million diabetes patients in Japan, and more than 1 million are on insulin therapy. Diabetes mellitus causes various complications (retinopathy, nephropathy, neuropathy, stroke, myocardial infarction, arteriosclerosis obliterans of the lower extremities, etc.), and the development of the complications significantly reduces QOL and life expectancy of the patients. Proper blood glucose control is essential to prevent complications, and therefore diet and exercise therapy and drug treatment (hypoglycemic agents and insulin) are important. It is important to mirror the physiological blood insulin levels of healthy people for adequate glycemic control of diabetics, and a typical means for this purpose is the intensive insulin therapy (multiple daily insulin injection regimen) with additional insulin and basal insulin supplementation. DM-SAiL supports non-diabetologists to perform intensive insulin therapy. DM-SAIL is the SaMD that provides recommended dosage of rapid-acting insulin and long-acting insulin for use with intensive insulin therapy.

We have completed the analysis of data from approximately 1,000 patients (about 1,080,000 clinical parameters) admitted to Tohoku University Hospital and have developed the AI that predicts insulin dosage with an error margin of about 2 units from the dosage prescribed by diabetologists. Currently, in collaboration with NES, we have been developing a system to utilize this AI in medical institutions and have completed the development of a demo system. In April 2022, the project was selected for the “Commercialization Promotion Project for Medical-engineering Collaboration (Support for small and medium-sized enterprises development and commercialization)” by AMED, and we will further develop the SaMD including clinical trials with the support of AMED. In April 2023, we filed an international application for the intellectual property.

Image of insulin action