Tal Arbel: Pioneering researcher challenges gender stereotypes

Posted on Thursday, November 16, 2017

Tal Arbel began programming basic video games before she hit her teens. Today, she’s a leading expert in computer vision.

By McGill Reporter Staff

When Tal Arbel was a kid, her father introduced her to computers and programming.

“I was 12 years old when my dad bought me a PC. It was a TRS-80. I learned to program basic video games (like Pong) on that machine. It had a tape recorder for storage. I was a teenager when I began to solder,” she says. “As an adult, however, people are often surprised when I tell them about my profession and say things like ‘You don’t look like an engineer!’ I am happy to challenge those stereotypes.”

Arbel has come a long way since those first forays in computer engineering, earning three Engineering degrees at McGill and becoming a full Professor in the Department of Electrical and Computer Engineering.

She’s also leading expert in computer vision, a sub-area of artificial intelligence (AI), focusing on a number of application areas in healthcare. Advanced techniques developed by Arbel’s group for detecting and analyzing lesions in brain images of people with Multiple Sclerosis (MS) are currently being used in an industrial software analysis system used for drug development for almost all MS clinical trials for new treatments worldwide.

“When I completed my PhD, I did a postdoctoral fellowship at the Montreal Neurological Institute where I began to realize that methods developed in computer vision have the capability to make concrete advancements in the areas of neurology and neurosurgery,” explains Arbel. “Through collaborations with neuroscientists, clinicians, biomedical engineering researchers and medical industrial partners, my research has been focused on developing machine learning frameworks to address a number of open problems in neurology and neurosurgery.”

“We are developing mathematical algorithms and computer code to get the computer to autonomously understand what it is looking at,” she continues. “Medical image analysis is an area within computer vision where the images are acquired from medical devices such as Magnetic Resonance machines, ultrasound, x-rays, etc. The methods developed in the lab have been used to improve the speed, accuracy and cost of an image analysis software pipeline used for new drug development in MS and will be used to assist surgeons in neurosurgical pre-operative planning.”

Arbel has built an internationally-renowned, multidisciplinary research program focussing on new probabilistic/mathematical frameworks in computer vision. Her research could lead to the automatic identification of biomarkers of progression in MS and analysis of brain tumours.

“McGill is an international leader in the areas of artificial intelligence, computer vision and neuroscience (including neurology and neurosurgery). Through interdisciplinary collaborations at McGill, my research team in the Faculty of Engineering is developing new techniques in machine learning, computer vision and brain imaging, and applying the techniques to help improve medical care for patients with diseases such as brain tumours and Multiple Sclerosis,” says Arbel. “Seeing this work begin to show potential impact in healthcare has been incredibly satisfying for me.”

No time for downtime

Arbel is a busy person. With an extensive list of publications to her credit, she also chairs and participates in many different scientific bodies at McGill, across Canada and internationally. Many of the graduate students whose work she oversees have gone on to break new ground and are heavily sought after by industry and academia.

On top of that, Arbel teaches a variety of large undergraduate courses in the department (some at enrolments of 100-150 students), mainly in computer engineering. She developed a new popular undergraduate course called Introduction to Computer Vision, and has redesigned a graduate course called Statistical Methods in Computer Vision that she also teaches.

Arbel says she is particularly proud of the interdisciplinary work being done at the Centre for Intelligent Machines (CIM), a research center focusing on several areas of intelligent systems that includes faculty and students from the School of Computer Science, Department of Electrical and Computer Engineering, and the Department of Mechanical Engineering.

At CIM, research is being done on robotics, artificial intelligence, computer vision, medical imaging, virtual and shared reality, control systems, computer animation and reinforcement learning.

Advocating for more women in science

As one of few women at the CIM and in Electrical and Computer Engineering, Arbel is working to promote science and engineering as fulfilling careers for women, within her department, faculty and within her research communities. Part of the problem she says, is that many women and girls aren’t necessarily aware of the wide range of possible careers in science and engineering, nor encouraged to enter these fields.

“I think the wide range of career opportunities afforded by having a degree(s) in engineering is not well known. Most of the engineers I know have a relative who is an engineer,” she says. “As engineering is still male-dominated, girls who are good at math and science are not generally encouraged to pursue careers in engineering, with the thought that it is more suited for males – leading to a vicious cycle. I find that when I speak to young women about my career, they are often surprised to hear that I am an engineer, and say ‘Had I known about a cool career such as yours, I would have gone into Engineering.’

Furthermore, there are very fewer role models and mentors for women interested in pursuing a career in academia in Computer Science and Engineering, and unconscious biases still exist.

Arbel is happy to be a role model.

“Outreach and information sessions about careers in engineering and in academia are required at the high school level and earlier,” she says. “Increasing the visibility of roles models will help to remove stereotypes and to inform the public about career opportunities these areas.”

Lessons learned at home last a lifetime

Arbel says that the most important lessons for young girls begin at home, citing both of her parents in laying the foundation for her own remarkable journey.

“My parents have been incredibly supportive of my career choices,” says Arbel. “My mom has always been a huge proponent of doing what you love and working hard to achieve your goals. She believes that women should be independent and has always tried to steer us towards careers that have a positive impact.

“My father and I have always had many interests in common. We have both always been fascinated by math and science,” she continues. “He exposed me to engineering from a very young age by explaining to me what it is that he did, by bringing home gadgets for us to play with (such as the personal computer), and by bringing home models planes and Lego kits for us to build together. He loved to talk to me about the math and science that I was doing in school. My father is indeed progressive, and I am sure that he treated me exactly as he would have should I have been born a male.”

Those early experiences at home had a huge positive impact on Arbel who, now with a family of her own, is sharing her love of science and technology with her children.

“My kids have shown great interest and aptitude in math and science so far, and enjoy playing video games. My older son (15) provides tech support for the family. My younger son (12) has been to robotics and to coding camps,” she says. “My husband is also a McGill Electrical Engineering graduate [Dan Wood, B.Eng. ’94]. He is now a Vice President at Intel Corporation, in charge of strategic planning for visual technologies. There is a lot of talk of technology at our house.”

Whether it is in the lab, in the classroom or at home, Tal Arbel is passing on the tradition of enquiry and curiosity.

 

Share this article

Category: Profile

Tag: , , , , , , ,

Post a Comment

  1. You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>