Accelerating AI-assisted Digital Health in research and beyond

Written by Dr Muhammad Fermi Pasha, School of Information Technology

Digital Health has become very popular nowadays and is rooted in various known areas in the past decades, such as eHealth, healthcare IT, medical informatics, and others. Simply put, digital health is the use of IT innovation in healthcare and its related fields. Undeniably, Artificial Intelligence (AI) is one of the drivers that assist the growth of Digital Health, and we have been fortunate to take a small part in this exciting research area. Some notable Digital Health technologies that we have been researching in the past include computer-assisted diagnosis (CAD) tool that uses various techniques such as 3D modelling and medical image analysis to assist clinicians in making the diagnosis, adopting clinical document architecture (CDA) as electronic health record (EHR) to transform paper-based record into electronic/digital record. Picture archiving and communication system (PACS) technology uses DICOM standards to act as a centralised platform for clinicians to access digital medical images regardless of modality.

Presently, we are currently making efforts to accelerate AI-assisted Digital Health research further on various fronts. Three Monash Malaysia funded PhD students working on the emerging field of Digital Health research area, namely: (1) a research project that aims to propose an analytical mapping framework to intelligently map electronic health records in older standards into the growingly popular fast health interoperability resources (FHIR) standard, (2) combining blockchain and AI to propose a novel and secure access based control framework for cross-domain personal health record (PHR), and (3) researching AI-assisted mHealth app as a self-help tool towards detection and prediction of depressive symptomatology amongst adolescents. Lastly, together with international and industry collaborators, our multidisciplinary team devises a profound convolutional networks architecture as a deep learning tool to understand the child's social learning behaviour by recognising their microexpression under the FRGS grant-funded research project. The research aims to provide psychotherapists with an AI-assisted digital health system in performing technology-enhanced therapy using historical videos with auto-tagged microexpression analysis.

Beyond research, the current ongoing pandemic certainly has accelerated digital health adoption by the healthcare industry worldwide. It would open up opportunities for university-industry collaboration. To name a few, the growth of both online remote doctor consultation and on-demand doctor house visit increases the need for centralised health record. Our existing research on FHIR-based analytical mapping framework and a blockchain-based access-based control framework could play a role in having a centralised patient-centric health record repository system where various legacy and non-standard health records can be migrated automatically to FHIR-based health records. Patients can give authorisation to their doctors to access their complete health records from anywhere. Proposals can be made on various analytical engines to support evidence-based medicine efforts. There is also an increased digitalisation effort in the healthcare industry to support the drive of what is known as consumer-centric digital health systems where health-conscious consumers are spending more on AI-assisted mobile apps with smart sensor IoT technology that allows them to monitor their health and get first-level remote help whenever they need medical advice from a doctor. We are open and looking forward to collaborating with any healthcare institution to accelerate their digital health adoption, especially with the help of AI technologies.