Speeding Up Examination Of Heart Diseases Through The Use Of 3D Echocardiography And Machine Learning

Dr Wong, Kenneth Kwan Yee   (Project Coordinator (PC))

Professor Lee Alex Pui Wai   (Co-Investigator)

Dr Choi Yi King   (Co-Investigator)

48

2021-04-01

5415350

Speeding Up Examination Of Heart Diseases Through The Use Of 3D Echocardiography And Machine Learning

3D Echocardiograph, Heart Diseases, Machine Learning

Others - Medicine, Dentistry and Health

Engineering (E)

MRP/027/20X

Midstream Research Programme for Universities (MRP)

2020

On-going

Heart diseases are the leading cause of death both globally and locally. The major diagnosis tool for heart diseases is 2D echocardiography, which is an ultrasound-based imaging technique. To perform an echo examination, it takes a trained specialist 30-45 minutes to capture multiple cross-section views of the heart by manual manipulation of a transducer. This lengthy process causes a long waiting time (an average of 81 weeks) for an echo examination in Hong Kong public hospitals, which inevitably delays the diagnosis and treatment of patients. The objective of this project is to speed up examination of heart diseases using 3D echocardiography. 3D echocardiography can acquire a volumetric representation of the heart within 10 seconds, but its visualization and interpretation are non-trivial. We will develop AI software using deep convolutional neural networks for automatic reconstruction of cross-section views of the heart from 3D echo data, and for automatic diagnosis of common structural heart diseases. We will train our deep neural networks using 3D and 2D echo data of local patients in the past 10 years. We target at reducing the examination time to less than 3 mins. We expect this can shorten the average waiting time for an echo examination from 81 weeks to below 10 weeks.