Prediction of stroke using ultrasound video of atherosclerotic carotid plaque
Frederick University & University of Cyprus
Advancements in Medical Image Computing and Artificial Intelligence (AI) have unleashed various applications in Medicine and Biomedical Engineering for the advancement of human health. One of these applications is the assessment of the risk of stroke in ultrasound video of atherosclerotic carotid plaque. An important international collaboration between the Frederick University, University of Cyprus and other universities abroad has been exploring this application.
As Dr Efthyvoulos Kyriacou, the director of the eHealth Laboratory, explains, the use of ultrasound for the prediction of stroke has been his a long-term effort for the past 15 years of his academic career. Early prediction of the disease could prove pivotal in treatment.
Cardiovascular disease (CVD) is the first leading cause of death and adult disability in the industrial world. According to Dr Kyriacou’s review publication (https://doi.org/10.1109/TITB.2010.2047649), 80 million American adults have one or more types of CVD, of whom about half are estimated to be aged 65 or older. Of all the deaths caused by CVD among adults aged 20 and older, an estimated 6 million are attributed to coronary heart disease and to stroke, with atherosclerosis, the build up of plaque in the arteries, as the underlying cause. Stroke, which is caused by the cut off of blood supply to the brain, accounts for about 1 for every 16 deaths in the United States. A recent study by the World Health Organization revealed that by 2015 almost 20 million people will die from CVDs, mainly from heart disease and stroke.
High-Resolution ultrasound has made the noninvasive visualisation of the carotid artery that transfers oxygenated blood to the head and neck possible at the bifurcation point where the carotid divides into two smaller arteries, and has thus been extensively used in the study of arterial wall changes. During the past 20 years, the introduction of computer-aided methods and image standardisation has improved the objective assessment of carotid plaque echogenicity and heterogeneity, and has largely replaced subjective (visual) assessment that had been criticised for its relatively poor reproducibility. In other words, machines are becoming more reliable than humans, especially with the use of Machine Learning (ML), a subfield of AI.
In the aforementioned publication, several ML techniques are applied, such as:
- Self-organising maps (SOM)
- Back-propagation of feed-forward neural networks
- Radial Basis Function (RBF) networks
- Probabilistic neural networks (PNN)
- Support Vector Machines (SVM)
- k-Nearest Neighbours (k-NN)
In addition, a brief survey of a number of classification studies and comment on the association between the extracted plaque characteristics and cerebrovascular symptoms is provided.
In the concluding remark, it is noted that it would be interesting to develop noninvasive, multimodality plaque-image analysis systems. The advancement of 3-D ultrasound will help these efforts. Basically, high-resolution 3-D ultrasound reconstructions will be much easier to fuse with 2-D slices from other modalities. To do this, we would need to register the geometric features of the 2-D slice to the 3-D volume or to use a mutual information registration method.
Finally, the impact of this application is essential to human health. This is an excellent example on how machines with the advancement of AI can contribute to the human health, decreasing the human error.
Dr Efthyvlous Kyriacou is currently an Associate Professor in the Department of Computer Science and Engineering of Frederick University, Cyprus. Undergraduate and Graduate studies at the department of Electrical & Computer Engineering of the National Technical University of Athens (NTUA) Greece (Diploma in Electrical and Computer Engineering 1996, Ph.D. 2000). His research interests focus on ehealth systems, emergency telemedicine systems, medical imaging systems and intelligent systems applications in medicine. He has published 37 refereed journal, 105 conference papers, 21 invited book chapters and has one patent in these areas. He has been involved in numerous projects in these areas funded by EU, the National Research Foundation of Cyprus, the INTERREG and other bodies. He serves as a reviewer in many journals related to his research fields. He was the Program Co-chair of ITAB 2009, BIBE 2012, Medicon 2016, Melecon 2016 and was in the program committee of several other scientific conferences. He is a Senior Member of the IEEE and currently is the chairman of the IEEE Cyprus Engineering in Medicine and Biology/Signal Processing chapter.
More information regarding the work at the eHealth Laboratory can be found at:
We are grateful to Dr Efthyvoulos Kyriacou for providing us with the relevant information needed to write this article. We would like to wish him all the best in the future.
