Researchers from A*STAR leverage high performance computing for cardiovascular disease stratification for early diagnosis and prevention.

Cardiovascular disease is the leading cause of death globally, taking an estimated 17.9 million lives each year and representing 32% of global deaths. Of these deaths, 85% were due to heart attack and stroke. In Singapore, cardiovascular disease accounts for 31.7% of all deaths in 2020 and 19 people die from heart diseases and stroke each day. This means that almost 1 out of 3 deaths in Singapore is due to heart diseases or stroke.

Over the last decade, technological developments have strengthened the role of non-invasive imaging for detection, risk stratification and management of patients with heart disease. There is a desire in the cardiac magnetic resonance imaging (CMR) community for consistency, precision and efficiency of CMR analysis. Machine learning has showed great promise in speeding up analysis which reduces the time spent on reporting by clinicians and removes intra and inter observer variability.

A team of researchers at A*STAR’s Clinical Data Analysis and Radiomics group are harnessing NSCC’s supercomputing resources to build a complete end-to-end deep learning pipeline for cardiovascular disease stratification for early diagnosis. The team has built a cardiovascular data processing pipeline which can handle large datasets of CMR/Cardiac-CT data from multi-centres. The data files are converted to a standard data structure format called cardiac data structure (CIDS). A model of different deep learning architectures are created which can be used to extract and quantify different cardiac measures like LV – volume, epicardium and endocardium volume and structural changes, ejection fractions and others for early diagnosis of heart disease. Work is in progress to make it into a deployable solution and to extend it for large cohort studies.

To find out more about the NSCC’s HPC resources and how you can tap on them, please contact [email protected].

NSCC NewsBytes April 2022

Other Case Studies

Designing safer nanomaterials and medicines with the aid of supercomputers

Researchers from Nanyang Technological University leverage high-performance computing resources to better understand the interactions between nanomaterials and biological...

Supercomputers help scientists better understand peanut allergies

Researchers from A*STAR leverage high-performance computing (HPC) resources to develop advanced structural model and simulations of the interactions between antigens and...

Leveraging machine learning and supercomputers to accelerate new materials development

Researchers from NUS utilise high performance computing resources to accelerate the discovery and development of new materials. The development of novel functional materials has...