Ishani DasGupta
2023 Runner Up
Computing Science, College of Natural and Applied Sciences
Listen to Your Heart; Let Machine Learning Do the Rest
Introduce yourself…
Hello! I am a Computing Science graduate student working towards my Ph.D. at the University of Alberta. I completed my M.Sc. in Electrical Engineering at the University of Calgary and moved up North. I am an introvert. I enjoy quotations, fandom references, plot twists and snowfall. As a Calvin & Hobbes fan, I am naturally drawn to adventure and am passionate about photography. When not out exploring, you will find me reading, cooking or cheering on my favourite football (soccer) teams.
What are you researching and what do you hope comes out of your research?
My work involves scrutinizing ultrasound images of the heart to assess its performance. I am working towards building a self-learning platform that can accurately identify the left ventricle (LV) and analyze heart function. By examining heart images captured at different stages of operation, my goal is to develop an automated system that can efficiently pinpoint any irregularities based on the shape and volume of the LV. To achieve this, I design and train neural networks that continuously learn from these images and mimic a human brain to identify potential issues with blood circulation and pumping action.
This training process deals with vast amounts of data and images. Working closely with cardiologists and radiologists at the University Hospital, I strive to improve the speed and accuracy of identifying cardiovascular diseases, a significant health concern for many people.
My ultimate objective is to develop efficient and reliable tools that can aid in diagnosing heart problems, ultimately enhancing the quality of life for countless individuals.
How did presenting a Three Minute Thesis (3MT) help explain your research to the public?
The project we are tackling requires the combined efforts of doctors and engineers. Condensing my research into a 180-second presentation, devoid of technical jargon yet focusing on the critical issue, is challenging.
The success of our collaboration hinges on my ability to communicate my ideas effectively and concisely to a diverse team of specialists.
Participating in the 3MT helped me to refine my communication skills and present my ideas in an engaging and impactful manner. Ultimately, it is all about bridging the gap between research and public understanding. The 3MT is an excellent platform for achieving that goal.
What inspires you to do research?
My parents are the most important people in my life. As professors themselves, their unique perspectives and advice have always guided me. Every bit of the person I am today is who they taught me to be.
What are three key words important to your 3MT?
Left ventricle, heart, neural networks
How does your research impact local, provincial, or global communities at large?
Cardiovascular diseases affect millions worldwide. According to the Canadian Chronic Disease Surveillance System (CCDSS), about 1 in 12 (or 2.6 million) Canadian adults aged 20 and over live with diagnosed heart disease. Our research project has the potential to revolutionize cardiovascular healthcare by improving patient outcomes and reducing wait times. We are developing an automated system that utilizes machine learning to identify the left ventricle (LV) from ultrasound images of the heart. This system aims to reduce the workload of doctors and medical personnel and also improve accuracy by reducing inter-and intra-observer variability and patient dependency. By identifying the LV and analyzing heart images, our system will enable experts to effectively respond to patients, even in remote locations, ensuring timely diagnosis and treatment.
If you had to dedicate your research to anyone from the past, present, or future, who would it be and why?
I dedicate my research to my family, halfway around the world, but always with me.
3MT Image Description: Sound waves moving through the heart (on the left). An arrow taking these sound waves to create echocardiography images on the right where the left ventricle (LV) is highlighted. The arrow has a neural network as a backdrop to indicate machine learning.