A new AI-based tool that gamifies the screening of mild cognitive impairment has been incorporated into a comprehensive health screening service in Singapore.

ReCOGnAIze is a neuroscientific game application developed at the Lee Kong Chian School of Medicine (LKCMedicine) of Nanyang Technology University, Singapore.

WHAT IT'S ABOUT

Powered by a proprietary algorithm, ReCOGnAIze comprises four games designed to assess different brain functions:

• a processing speed game where numbers are matched with symbols; 

• an executive function game where users create a trail of numbers and symbols;

• an attention game where aeroplanes go in alternate directions and colours; and 

• a working memory game where items from a grocery shopping list must be memorised and recalled.

These games, developed over three years at LKCMedicine's Dementia Research Centre, can detect early signs of cognitive impairment in 15 minutes. 

"Gamification enables us to pick up different features or metrics from the gameplay and helps in detecting specific brain functions which could be related to cognitive decline," explained Mohammed Adnan Azam, co-founder of Gray Matter Solutions, an NTU Singapore spinoff company which owns the license to ReCOGnAIze. 

Besides being engaging and interactive for users, "the interactions and features captured from gamification provide a lot more in-depth insights to enable in-depth machine learning and deep learning models," he added.

The app has been validated against biomarkers and brain images of 230 individuals in the Biomarkers and Cognition Study, Singapore. It demonstrated 89% accuracy in identifying MCI. 

WHY IT MATTERS

Over a tenth of people with MCI progress to dementia each year worldwide; close to 10 million are diagnosed annually, according to the World Health Organization. 

In Singapore, around 86,000 individuals are living with dementia as of 2021, with the number projected to grow to 150,000 by the end of the decade. 

Dementia among Asians is commonly driven by vascular changes in the brain, such as silent strokes, according to findings from LKCMedicine research. That is why early detection of cognitive impairment requires more than just measuring memory, NTU Singapore stressed. 

MCI is traditionally detected and confirmed through neuropsychological evaluations and imaging tests. An AI-powered gamified screening alternative provides a "scalable and cost-effective way to detect MCI, allowing timely medical interventions that can significantly improve quality of life," noted NTU Singapore associate professor and Gray Matter Solutions co-founder Nagaendran Kandiah. He also heads the LKCMedicine Dementia Research Centre. 

Gray Matter Solutions recently partnered with international private GP clinic group, Osler Health, to launch its MCI screening app; the latter will offer the tablet-based app for free in the next three months as part of a holistic health assessment package for patients. 

"We have had some conversations with other groups as well. We're keen to have more such conversations, however, we would like to only work with those who are aligned with the vision of our company," Azam said when asked by Mobihealth News about their partnership plans. The company looks to expand their range of products and services, cater to mass public screenings, and partner with health organisations locally and abroad.

For its future commercialisation, Gray Matter Solutions is currently going through its first round of fundraising. "We expect to close it in three or four months, targeting between $1 million and $1.5 million," Azam shared.

MARKET SNAPSHOT

Another innovation in Singapore gamifies the communication training for caregivers of people with dementia. The Institute of Mental Health worked with Taggle to develop the Play2Care app, which incorporates the institute's standard approach to communicating with people with dementia. 

AI has been increasingly applied in MCI screening over the past years across Asia. In 2021, Singaporean private tertiary care provider Farrer Park Hospital rolled out an AI tool that measures brain tissue volume in the MRI scans of people suspected of dementia. Researchers at Korea's Electronics and Telecommunications Research Institute recently developed a large language model-powered tool for detecting MCI. Japanese pharma company Eisai reportedly developed a similar MCI risk prediction algorithm.