Age: 17Sunnyvale, CA
Project Title: MapAF: Deep Learning to Improve Therapy of Complex Human Heart Rhythm Abnormalities
Natasha M. Maniar developed a computational approach to identify sources of atrial fibrillation (AF). Despite affecting more than 33 million people worldwide, diagnostic imaging of electrical conduction through the heart remains relatively subjective and continues to rely heavily on visual interpretation by experts. Natasha addressed this as a two-fold problem. She first developed an algorithm to analyze the heart’s chaotic electrical signals and then interpreted those results using her computational tool. Her code identified the AF sources inside the heart with greater accuracy than trained experts. This tool improves AF treatment by streamlining and standardizing the catheter ablation procedure, making it globally accessible.
I’m Natasha, a 17 year old student from the Bay Area of California and incoming freshman at MIT. As an avid computer science enthusiast, I’m passionate about the applications of computer science to healthcare. I’m extremely grateful and honored to be named a Davidson Fellow. I’ve always looked up to previous Davidson scholars, and it’s immensely gratifying to be recognized for my work. I’m excited to join the Davidson Fellows community among many inspiring people.
For my project, I developed a computational tool to automatically detect source points of arrhythmia. Arrhythmia (heart rhythm disorder) is among the leading causes of death worldwide. Affecting over 30 million people, atrial fibrillation (AF), the most prevalent and complex arrhythmia, has serious, possibly fatal health effects. However, due to the irregular electrical activity and chaotic nature of AF in the heart, treatment options are poor and patients often require multiple procedures to cure the condition. Recent studies that have used voltage mapping reveal areas of rotational sources in the atria, for which ablation (burning of diseased tissues) at these particular regions completely terminates AF. Unfortunately, current methods to identify AF sources from these complex mapping videos are solely manual, therefore limited and subjective. I developed MapAF, the first computational approach using convolutional neural networks to automatically recognize the location of these AF rotational sources from within chaotic electrical patterns and improve AF therapy. Working on research projects in the electrophysiology field combining computer science and cardiology has been one of my most meaningful high school experiences—both because of the intellectual challenge and the personal ramifications. My research has allowed me to address the history of heart disease in my own family and many other patients suffering from AF, being one of the most prevalent and complex heart disease.
This project wouldn’t have been possible without the guidance of my mentors in Stanford Computational Arrhythmia Lab and The Harker School. I’m lucky to have such a close-knit, supportive, and intelligent lab group to work with. Dr. Sanjiv Narayan has been my biggest inspiration and role-model and has given me the confidence to believe in my work throughout my research journey. I’m also very grateful for Mahmood Alhusseini and Firas Abuzaid for their invaluable feedback, assistance with the computational pipeline, and providing the binary labels for the dataset. Ms. Anita Chetty, chair of the science department, and my science teachers, Mr. Mike Pistacchi and Mr. Chris Spenner have helped me think about the broader perspectives of my project and influenced my development as a scientist and altruistic individual.
MapAF is the first computational tool ever developed on the current, most accurate and effective method of ablation (rotational mapping). It will enable doctors around the world to provide better treatment for AF patients due to the standardized and accurate deep learning algorithm, which precisely locates the patient’s AF source(s) with 95% accuracy. The computational approach could save a significant amount of time during ablation procedures by eliminating long interobserver analyses and simply maintaining expert confirmation. It may also reduce variations of therapy—enable new laboratories to achieve similar results compared to those of highly experienced laboratories. Notably, this model will improve therapy for AF and may shed light on other disease causes not visible to the naked eye.
Since sixth grade, I’ve had the opportunity to dabble in a breadth of different disciplines that Harker offers. Some of the most interesting conversations arose in classes like Literature into Film and Advanced Research—these courses have broadened my horizons and taught me invaluable skills. Being in a STEM school, I’ve also had the chance to delve deep into advanced computer science and math classes such as Compilers, Expert Systems, and Linear Algebra. Aside from Harker, I’ve had the opportunity to participate in the Stanford Institute of Medical Research (SIMR) Bioengineering internship which has sparked my interest in interdisciplinary studies. I’m currently undecided on my major, but I’m planning to focus on using computer science applied to healthcare at MIT. I’m thrilled to learn from accredited professors and have access to incredible facilities and resources.
The capability of computational biology to solve major health issues and the potential of automated, computer based systems to make treatment more efficient and accessible fascinates me. Working on these projects allowed me to extend beyond the confines of my high school curriculum and make an impact in the real world.
Over the past couple years, I’ve also developed an iPhone app (MapAF) to train physicians on identifying sites for AF ablation therapy and was invited as an oral presenter to Heart Rhythm Society Conference 2018 to present my work. My proudest accomplishment was being named a Top 40 Regeneron STS finalist in March 2019 for an earlier version of the work I submitted for the Davidson Scholarship. Besides research, I was the founder and co-president of the Artificial Intelligence Club at my school. I’m also a nationally competitive dancer and have been dancing for 14 years. Dance has always been my creative outlet and an inspiration for many projects and ideas. Through my research experience, I have become more interested in the medical applications of artificial intelligence (AI), and I hope to someday spearhead a biotech company.
Where do you see yourself in 10 years?
Working in the intersection of computer science and healthcare to combat some of the biggest health issues.
If you could have dinner with the five most interesting people in the world, living or dead, who would they be?
Sheryl Sandberg, Stephen Hawking, Elon Musk, Martha Graham, Michelangelo
If you could be on any TV show, which one would it be?
So You Think You Can Dance
In the News
Natasha Maniar to be awarded $50,000 as a 2019 Davidson Fellow Scholarship Winner
San Jose, Calif. – The Davidson Institute for Talent Development has announced the 2019 Davidson Fellows Scholarship winners. Among the honorees are Cynthia Chen, 17, of Cupertino; Natasha Maniar, 17, and Aryia Dattamajumdar, 17, of Sunnyvale; and Anna Quinlan, 18, of Atherton. Only 20 students across the country are recognized as scholarship winners each year.
“I’m extremely grateful and honored to be named a Davidson Fellow,” said Maniar who will be attending Massachusetts Institute of Technology in the fall. “I’ve always looked up to previous Davidson scholarship winner, and it’s immensely gratifying to be recognized for my work. I’m excited to join the Davidson Fellows community among many inspiring people.”
Maniar developed a computational approach to identify sources of atrial fibrillation (AF). Despite affecting more than 33 million people worldwide, diagnostic imaging of electrical conduction through the heart remains relatively subjective and continues to rely heavily on visual interpretation by experts. Her code identified the AF sources inside the heart with greater accuracy than trained experts.
Click here to download the full press release
Click the image to download hi-res photos of Natasha:
Started in 1999, the Davidson Institute for Talent Development is a 501(c)3 private operating foundation. Our mission is to recognize, nurture and support profoundly intelligent young people ages 18 and under, and to provide opportunities for them to develop their talents to make a positive difference.
Profoundly gifted students are those who score in the 99.9th percentile on IQ and achievement tests. Read more about this population in this article.