Digital Healthcare and MedTech

Background / Motivation

Health and Medical Technology is an important sector in Singapore that supports more than 9,000 jobs with 50 Regional Headquarters, 25 MNCs with R&D presence, over 250 MedTech start-ups and small-medium enterprises3. Increasingly, new approaches to innovation in this sector will require sustained R&D efforts to help medtech companies move up the value chain. The COVID-19 pandemic has also accelerated the shift of healthcare systems to the digitalised paradigm and is rapidly transforming how healthcare professionals and stakeholders will work together for enhanced resilience.


To deliver scientific and technological advancements that impact the Health and Medical Technology sector, IHPC has set up the Health and Medical Technology Innovation Technology Area (ITA) group to harness the strengths of multiple disciplines across different departments.  

IHPC’s digital healthcare and medtech efforts include: 

  • Digital healthcare, especially in ophthalmology
  • Health-tech for wellness and pre-disease screening
  • Geriatric care
  • Modelling and simulation for production and design of medical devices 

Digital Healthcare

Multimodal AI for Ophthalmology

Multimodal AI for Ophthalmology
Fig 1: Detection of eye diseases using multimodal AI

Working closely with the clinical team from the Singapore Eye Research Institute (SERI) under SingHealth, the IHPC team has developed AI algorithms for eye disease detection (Fig 1). These include Diabetic Macular Edema detection, glaucoma detection from optical coherence tomography (OCT) scans, papilledema detection from fundus photos, and myopic macular degeneration from fundus photos. Some representative publications include Papilledema Detection (New England Journal of Medicine 2020), Diabetic Macular Edema (DME) detection (MICCAI2019), Retinal Vessel Segmentation (OMIA, MICCAI2019). The eye is also the window for human health, and the team has obtained promising preliminary results on cardiovascular disease detection and chronic kidney disease using multimodal data including fundus photos, clinical risk factors, genetic data and OCT.

A key highlight was the AI system developed by IHPC that can identify optic nerve abnormality (papilledema) with high accuracy. It can alert doctors of the possibility of a severe brain condition and minimise the risk of overlook during an emergency assessment. The performance of this deep learning system at classifying optic disc abnormalities was comparable to having two expert neuro-ophthalmologists. The study was published in the New England Journal of Medicine (NEJM) in April 2020 and the Annals of Neurology in July 2020 and also was reported by The Straits Times.

Health-tech for Wellness and Pre-Disease Screening

Doctor Covid Chatbot

The Doctor Covid Chatbot (Fig 2) was jointly developed by SingHealth1 and IHPC with support from the Integrated Health Information Systems (IHiS) to improve communication with Covid-19 patients while minimising transmission risk among healthcare workers. Patients would receive daily broadcast messages such as reminders and other medical information, as well as regular check-ins on their mental well-being by subscribing to the chatbot service. Doctor Covid has been registered by 3,000 patients staying in community care facilities (three locations in Singapore Expo / D’Resort NUTC chalet / Bright Vision Hospital). The system has been reported by The Straits Times and Lianhe Zaobao.

Doctor COVID Chatbot
Fig 2: Screenshot of Doctor Covid on Telegram

Efficient Interpretation of Chest-X ray Images

A deep learning algorithm was co-developed by A*STAR’s researchers and radiologists from Tan Tock Seng Hospital (TTSH) for detection of pneumonia from Chest X-ray (CXRs) images to efficiently interpret patient’s CXRs with a higher probability of pneumonia for priority screening (Fig 3). The developed CXR AI tool (now known as RagiLogic) was deployed at TTSH for the clinical team to conduct an independent evaluation on their existing workflow (Fig 4). The tool has helped to reduce the number of false-positive cases of pneumonia and has an accuracy of up to 96.1 per cent.

Chest X-ray screening of pneumonia
Fig 3: The proposed workflow for using deep learning for Chest X-ray screening of pneumonia

Dr Ting Yong Han (TTSH)
Fig 4: Dr Ting Yong Han, consultant in diagnostic radiology (clinical) at TTSH, looking at a chest X-ray
of a Covid-19 patient that was flagged by the AI tool, RadiLogic, which can analyse an image
within three seconds
Photo credit: The Straits Times

Geriatric Care

Dementia: Caregiving, Decision Support, and Pre-Screening

While caregiving for dementia patients is long-term and needs dementia-specific coping knowledge, healthcare technology is crucial in supporting caregivers to better care for their loved one with dementia. Such technology attempts to improve the wellness of both dementia patients and their caregivers by empowering caregivers with personalised intervention strategies and coping skills in care delivery. Technology applications for pre-dementia screening help people-at-risk to identify risk factors and prevent the development of dementia by recommending behavioural nudges and healthier lifestyles.

IHPC’s HealthTech efforts for dementia include caregiver/patient support, knowledge management, psychological-model-informed community care model design, and personalised behavioural intervention recommendations. By leveraging on a knowledge-based decision support system (Fig 5) with risk factor assessment, and psychological modelling, both the wellbeing of patients as well as their caregivers, could be better managed (e.g., enhancing emotional/physical health, stress management, promoting social engagement and healthy lifestyles). 

Application areas include:

  • Dementia management in community or at home
  • Personalised decision support for dementia caregiving
  • Caregiver stress management and coping strategies
  • Dementia patient engagement and leisure activities
  • Collaborative dementia caregiving
  • Dementia service provider recommendation
  • Pre-dementia screening for healthier lifestyles and behavioural nudges

Knowledge-based decision support system
Fig 5: Knowledge-based decision support system for dementia patients and caregivers

Cognitive Health Monitoring through Games 

There has been an increased focus on “serious games” that displayed positive effects on the cognitive abilities of elderly (e.g., older adults aged 65 and above) which could allow them to monitor their health independently. Yet there were very few games specifically targeted at the elderly population. 

Designing such games poses distinctive challenges since designers must take account of the potential cognitive, sensory, physical limitations of the elders, as well as their limited experience with gameplay interfaces and conventions compared to the younger generation. Cognitive abilities vary greatly among elders, especially among populations with a potential risk or already diagnosed with early-stage dementia. Hence, the purpose of each game should focus on different cognitive abilities targeting specific subgroups of the elderly population.

Towards this end, IHPC’s efforts in this area include the development of game design principles that involved cognitive abilities implicated at different stages of (normal and abnormal) cognitive ageing, and how these abilities would influence gameplay mechanics. With the effective design (depending on the type of game), the caretaker or medical professional could glean insights on users’ cognitive functions, such as executive functioning, prospective and working memory, hand-eye coordination and visuo-spatial tracking (Fig 6 & 7). 

Fig 6: Townlife: Find and deliver items by solving simple puzzles in an appropriate sequence

Fig 7: Trains: Fix misaligned tracks across a land area to ensure trains reach destinations

• Fua, K.C., Gupta, S., Pautler, D., & Farber, I. (2013). Designing serious games for elders. FDG.
• Farber, I., Fua, K.C., Gupta, S., & Pautler, D. (2016). MoCHA: Designing Games to Monitor Cognitive Health in Elders at Risk for Alzheimer's Disease. ACE2016.

Motor-Rehabilitation Enhancement with Rhythmic Entrainment  

IHPC has developed a motor-rehabilitation game system, known as MoMu (Move with Music, Fig 8), which leveraged the empirically-established facilitatory effects of rhythmic entrainment to promote motor rehab in stroke patients (especially balance, reaching movements) and strengthening in elderly. Rhythmic entrainment is a phenomenon through which a person’s bodily rhythms (e.g., body movements) would take the cue of and synchronise with an externally perceived rhythm, such as human music. Each time you hear a catchy tune and find yourself bobbing around to the beats of the music, that’s rhythmic entrainment at work.

Besides rhythmic entrainment, MoMu was also designed to track the patients’ progress and performance. The added benefit of music in motivating and promoting the enjoyment of rehabilitative exercises had made it suitable for elderly use. Healthy participants on trial with MoMu have performed with encouraging success, and hence on-going efforts towards clinical trials are underway.

Move to the Music (MoMu)
Fig 8: MoMu: A music-based motor rehabilitation game system in which motor rehabilitation & strengthening exercises are gamified with musical rhythms to promote recovery

Modelling and Simulation for Design of Medical Devices

Design of Medical Devices

Modelling and simulation play a critical role in the design, development, production, and clinical trials of medical devices. In some cases, modelling and simulation is the only way to evaluate device performance and often shortens the design cycle time significantly. Modelling and simulation are often referred to as an in silico approach, and together with in vitro or in vivo testing, these form the foundational aspect for medical device development. 

IHPC’s expertise in modelling and simulation capabilities for medical device development include innovative design concepts, performance evaluation, and design optimisation. By leveraging tools in computational mechanics and design methodology, important aspects such as device performance and material failure behaviour can be well studied before physical prototyping and trial. 

Application areas include:

  • Carotid artery stent (Fig 9)
  • Mitral valve replacement
  • Aortic stent-graft
  • Shoe mechanics
  • Blood vessel patency
  • Bone fracture healing
  • Knee spacer/replacement
  • Eye biomechanical modelling

Carotid artery stent
Fig 9: Covered stent, mitral valved stent, and prostate biopsy (top from left to right). The corresponding simulation results are at the bottom for wall shear stress of artery, crimpability, dynamic deformation of prostate, respectively

Collaboration Opportunities 

Leverage the span of technical expertise at IHPC for technology advancement to help your business move up the value chain. 

We welcome interested party to collaborate in research and development relating to Digital Healthcare and MedTech in identifying new technology areas to develop new applications, multi-disciplinary projects or commercialisation outcomes. 

For more info or collaboration opportunities, please write to