Healthcare (Commonwealth Union) – A three-year research study and pilot program called SoundKeepers is bringing together seven partners from the healthcare and social sectors to develop an AI (artificial intelligence) tool that uses voice biomarkers to identify early signs of depression in seniors. Additionally, the initiative will establish a community-based intervention program to support managing the condition outside traditional healthcare settings.
Much like how blood and stool samples are gathered to inform doctors about a patient’s physical health, the voice biomarker tool will collect voice samples (with consent) to provide insights into a patient’s mental health. Once completed and if proven effective, the tool and program aim to equip professionals and patients with an objective way to foster open conversations about a topic that often remains hidden, challenging to articulate, and sensitive to discuss.
The seven partners collaborating on SoundKeepers include Nanyang Technological University, Singapore’s (NTU Singapore) Lee Kong Chian School of Medicine (LKCMedicine) and College of Computing and Data Science (CCDS); National Healthcare Group Polyclinics and Institute of Mental Health from the National Healthcare Group; social service organizations Fei Yue Community Services and Club HEAL; and philanthropic organization Lien Foundation. Support will also come from 20 GP clinics in Hougang and Woodlands.
The project, which involves over 600 seniors, seeks to assist those aged 55 and above experiencing subsyndromal depression (SSD), an early stage where depressive symptoms emerge but are not yet severe enough for a diagnosis. Among seniors, SSD poses a largely overlooked health risk.
Seniors with SSD are five times more likely to develop full depression within a year and are 12 times more at risk for dementia. In Singapore, SSD is one of the most prevalent mental health issues affecting older adults, with 13.4% of community-dwelling seniors over 60 reportedly impacted—a figure likely underestimated due to reliance on self-reported data. Seniors with SSD face higher healthcare costs than those without and show outpatient service usage similar to individuals with diagnosed depression. Conditions commonly affecting seniors, such as chronic diseases and disabilities, often worsen SSD.
Subsyndromal depression (SSD) is significantly more common than full depression, occurring three times as frequently and with an incidence rate five times higher. This means that at any given moment, there are three times as many individuals experiencing SSD as there are with diagnosed depression. Over a one-year period, the number of new SSD cases is five times that of new depression cases. The gap between prevalence and incidence rates is due to factors like recovery and mortality.
“We need new ways to listen to our seniors. While they may not express their worries through words, we can now try to hear it through their voices,” said Mr Lee Poh Wah, CEO, Lien Foundation.
“Currently SSD is not actively diagnosed or treated. However, with the focus on early detection and treatment emphasised by both HealthierSG and the National Mental Health and WellBeing Strategy, this project becomes extremely relevant as it can facilitate the early detection and diagnosis of SSD with a tool that can be easily used in the community setting,” said Dr Mythily Subramaniam, Assistant Chairman, Medical Board (Research), IMH and CoPrincipal Investigator of SoundKeepers.
Participant recruitment will take place at Hougang and Woodlands Polyclinics. The program will have two main parts: collecting voice samples at the polyclinics and the Institute of Mental Health (IMH) for the development of the voice biomarker tool at the College of Computing and Data Science (CCDS), and a referral process to connect identified seniors with social service agencies (SSAs) where they will participate in a 24-week early intervention program within the community.
In the first part, researchers will gather a voice sample, lasting several minutes, from each consenting participant through a casual conversation or a reading exercise. These samples will contribute to building the SSD-detection algorithm using a dedicated smartphone. Once created, the voice production biomarker will offer an objective measure of a patient’s mental health status, specifically around depression or subsyndromal depression, functioning similarly to how doctors currently use temperature or blood pressure readings.
