FRONTEO Aims to Develop the World's First Linguistic AI Medical Device

Expectations for Diagnose Dementia from Daily Conversations by Utilizing Conversational Dementia Diagnostic Support AI Program

Did you know that the Ministry of Health, Labor and Welfare predicts that the number of dementia patients over the age of 65 in Japan will reach about 7 million by 2025? This means that about one in five elderly people will have dementia. Dementia is one of the most difficult diseases to diagnose because it requires specialized knowledge and experience. However, the diagnosis of dementia will be dramatically changed to be easier with an AI medical device that is being developed by FRONTEO Inc. The AI medical device, which uses natural language, is expected to enable screening for dementia through a ten-minute conversation between a doctor and a patient. In this article, we will discuss FRONTEO's strategic program, which aims to develop the world's first linguistic AI medical device from Japan.

Dementia
Dementia is a disease in which cognitive functions deteriorate due to various causes such as brain diseases and disorders, causing problems in daily life. The diagnosis of dementia is based on interviews with doctors, neuropsychological tests, and diagnostic imaging. It is also said that it is difficult for doctors to diagnose dementia unless they are specialists. The challenge is that diagnosis takes time, which is a burden for both doctors and patients.

Conversational dementia diagnostic Support AI program
FRONTEO is developing an AI medical device using its AI engine “Concept Encoder.” It is expected that the AI program using natural language will enable screening for cognitive dysfunction through a ten-minute conversation between a doctor and a patient.

Concept Encoder
Concept Encoder is an artificial intelligence (AI) focusing on natural language developed specifically for the life science field. It was developed with the goal of effectively analyzing and utilizing life science-related big data based on evidence. Statistical methods such as significance tests, which are essential for “evidence-based medicine (EBM),” a common understanding among life science researchers and healthcare professionals, have been introduced and implemented in natural language analysis. It is also possible to perform co-analysis even with non-text data. The company is able to conduct research and development on co-analysis with "numerical data" such as gene expression information, vital signs, and various test values that have been accumulated in the life science field.

AI programs for dementia in clinical practice
In order to use AI programs in clinical practice, it is necessary to conduct clinical study (clinical trial) under PMDA (Pharmaceuticals and Medical Devices Agency) *¹ consultation. The clinical trial will compare a judgment by the AI program with one by the usual way of diagnosis to see if there is any difference. The way to use the AI program is very simple: the AI program analyzes the natural conversation between a doctor or psychologist and a patient in a usual medical visit with no additional works to support the doctor's diagnosis.

On the other hand, neuropsychological tests, one of the current methods of diagnosing dementia, can only be determined by a doctor with expertise. The results may vary depending on the doctor who evaluates the patient. The test also takes time. Therefore, this test is a burden not only on the patient but also on the doctor. The results of clinical trials are attracting attention to see if AI programs can assist doctors in diagnosis who address this burdensome challenge.

This program from the perspective of clinical development personnel
The process for FRONTEO to begin clinical trial must have been extremely tough and arduous.

(Continue to the next)

The information contained in this article is compiled by the respective authors based on publicly available information. We assume no responsibility whatsoever for any damage or disadvantage caused by actions taken based on such information. Unauthorized reproduction of articles, photos, charts, etc. is prohibited.
Copyright © 2022 LSMIP office / CM Plus Corporation