A team of Republic Polytechnic graduates has developed “Stroke Guard,” an AI-powered camera system designed to detect facial drooping in bedridden patients. By monitoring physical markers of stroke, the technology aims to reduce the reliance on 24-hour manual supervision in nursing homes. The project, led by Ashwin Manikandan, Tan Jia Ling Jolin, Sasindren Jaya Sanger, and Sarah Beegam Saju, received $3,000 in seed funding through the institution’s Entrepreneurship Immersion Programme to develop their minimum viable product.
How Does AI-Driven Stroke Detection Work?
Stroke Guard utilizes stationary cameras mounted at a patient’s bedside to track facial movements. According to the development team, the system identifies facial asymmetry—a primary indicator of a stroke—and automatically triggers an alert to caregivers. Unlike wearable medical devices that require physical contact with the patient, this non-invasive approach allows for monitoring without disrupting a patient’s rest. Sasindren Jaya Sanger, the team’s technical lead, noted that the software was built using Python and HTML, with the AI model refined through datasets collected in collaboration with Care Corner’s Active Ageing Centres.
What Are the Challenges of Scaling Medical Startups?
While technical development often dominates the early stages of a startup, the team identified business operations as their most significant hurdle. Tan Jia Ling Jolin reported that securing partnerships and navigating the medical technology market required persistent outreach. By engaging with mentors and proactively contacting organizations like Care Corner, the team expanded their reach. Following their presentation at the 2025 Danang Venture and Angel Summit, the group received interest from international investors, highlighting the global demand for scalable eldercare solutions.
Why Is Non-Invasive Monitoring Essential for Caregiving?
The push for AI-assisted care stems from the physical and mental strain placed on caregivers. Tan, who served as a full-time caregiver for her brother, emphasized that technology can alleviate the burden of round-the-clock monitoring. Current industry trends show a shift toward “passive” monitoring, where systems operate in the background. This contrasts with traditional medical hardware that often requires frequent battery charging, skin-contact sensors, or intrusive cables that can cause discomfort for elderly patients in long-term care facilities.
Frequently Asked Questions
What is the primary function of the Stroke Guard system?
Stroke Guard uses AI-powered cameras to detect facial drooping in bedridden patients and alerts caregivers to perform further medical assessments.
Does the system require physical contact with the patient?
No. According to Sarah Beegam Saju, the system is a camera-based, non-invasive solution that does not require wearables or physical attachments.
Who can benefit from this technology?
The system is currently designed for bedridden patients in nursing homes or similar care facilities where constant physical monitoring is difficult to maintain.
How was the AI model trained?
The team collaborated with Care Corner’s Active Ageing Centres, collecting over 70 datasets of facial expressions to improve the system’s accuracy and minimize false positives.
Are you interested in the future of medical technology and aging-in-place solutions? Subscribe to our newsletter for the latest updates on health-tech startups and industry trends. Join the conversation in the comments below—how do you see AI changing the landscape of eldercare in the next decade?
