Computer Vision-based Healthcare Monitoring
Dr Mohamed Hisham Jaward Lecturer PhD
Research expertise : Facial Expression Recognition , Activity Recognition , Video Tracking and Deep Learningbased Approaches to Biomedical Images
E : mohamed . hisham @ monash . edu T : + 603 5514 6204
Cardiac Arrhythmias : Understanding and Controlling
Dr Liang , Shiuan-Ni Lecturer PhD
Research expertise : Biophysics , Biomedical Engineering , Nonlinear Physics
E : liang . shiuan-ni @ monash . edu T : + 603 5514 4991
With the continuous improvement in living environment and healthcare , the population , in general , has a longer lifespan compared to a few decades ago . As a result , nursing homes around the world start to face a growing labour shortage . One of the solutions to this problem is the video-based technology for ambient assisted living . Similar techniques are also useful for monitoring the condition of a bed-ridden patient . In addition to applications in healthcare monitoring , automating facial expression analysis could promote facial expressions as part of Human-Computer Interaction ( HCI ) and make the interaction more natural and more efficient .
Cardiac arrhythmia is a condition where heartbeats are abnormal . Every year , millions of people worldwide experience arrhythmias and many of them can return to their normal lives if the arrhythmias are treated at early stages and properly . However , many arrhythmias are dangerous and may lead to sudden cardiac death . In a healthy heart , when the cardiac cells are stimulated by electrical impulses , action potentials can be initiated and propagate as electrical waves through the heart . At normal heart rhythm , the action potential duration ( APD ) is almost identical from beat to beat . When pacing becomes faster , cardiac alternans
Our current plans focus on addressing related problems using computer vision techniques . Specifically , two problems are being investigated : Facial expression recognition and human activity recognition . Our current research focus would be to investigate the deep learning algorithms for the following two problems : Osteoarthritis cartilage segmentation and palm seed classification . Our future research objective is to develop deep learning algorithms for bio-medical image processing problems .
can be developed . Cardiac alternans is clinically important because prolonged untreated alternans may lead to ventricular fibrillation ( VF ), a lethal arrhythmia which is the leading cause of sudden cardiac death . Our research focuses on understanding the mechanism of these cardiac arrhythmias , including alternans , ventricular tachycardia and VF , by using mathematical models and finding effective ways of controlling them . These projects are carried out in collaboration with partners from Taiwan and Vietnam .
One of models developed to model facial expressions
Segmented image of a knee cartilage
APD alternans before control ( first four beats ) and after control ( last four beats )
Time sequence of APD with APD alternans is successfully controlled at beat number 741
37ELECTRICAL AND COMPUTER SYSTEMS ENGINEERING
R E S E A R C H E R P R O F I L E 2022 / 2023