Interview with Ganesh Kumar

It was a privilege to interview Mr. Ganesh Kumar M.R., Software Architect at Applied Materials, who has over 10 years of experience in development of Image Processing Algorithms for medical and semiconductor products. His experience as a Junior Research Fellow (JRF) in IISc, Bangalore and over 9 years of experience in GE Healthcare, a company involved in medical imaging, made us inquisitive so as to know how the Medical Imaging Industry works!

Thank you, sir, for having agreed to do this interview! To begin with, you had done most of your education in Physics. What is it that made you so passionate about the field?

Physics is about trying to come up with explanations of nature-its origin and what it is- by equations and concepts. For example, when it rains, even the pattern of fall of the rain drops follows Poisson’s noise distribution. Similarly, rainbow formation too has physics involved behind it. This has intrigued me a lot and made me develop interest in it.

So, what made you switch to the Biomedical Field?

Even the biomedical field has a lot of physics involved, especially in developing medical equipment. There’s nothing that gives greater joy than knowing that whatever work you’re doing is going to save someone’s life. So that was the greatest motivation for me. Applying what I had learnt in Physics, especially X Ray Physics and Signal Processing as theory, enabled me to develop products with great care.

How was the experience of being in a reputed institute like IISc, especially as a JRF?

I would say that that was the first journey for me out of college, i.e. the first exposure to the R&D environment. It gave me a chance to learn on how to apply the concepts I had learnt to go deeper into an area. I had a chance to work on Raman Spectroscopy. I had to develop a tool for Atomic Force Microscopy. The project was quite complicated. Though we had funding, it took us a lot of time to get the equipments especially since most of the equipments were not available in India. Once, we had all the equipments it was more like designing an electronic circuit, to move the stages and aligning the images onto the cantilever of the Atomic Force Microscope. Getting Raman Spectroscopy from the tip of the AFM was a real challenge. It took us a nine-month effort to see the first image out of it. It was a great privilege because, many of us work with microscopes, but very few would get a chance for building one.

Regarding your work in GE Healthcare, which was the most challenging project you were given?

There were a lot of challenges right from day one. Among them, the most memorable one for me is trying to build a ‘Heart Simulator’. We were developing an Angiography equipment, for which we needed to simulate the beating heart. The challenge was to see how well the system could capture the beating heart, because the beating heart keeps moving and blurs the images. To do this, first we needed the data, which we were able to get from our legacy products on the field. But with this new Angiography equipment, it’s tough to get it into the field and get data. So, for this I needed to build the Heart simulator.

So, you tried to simulate the beating heart, so that while taking the angiogram, the noise due to its movement can be removed. Is that the crux of it?

You’re almost there! Even if you have a normal mobile phone camera, imagine ****trying to take an image in the dark. When the subject moves, you’ll see that the image is blurred right? In an angiogram similarly, X-rays are used and regulations demand that we need to give minimal X ray radiation dose to the patient. This is comparable to the low light conditions I explained above. And here the heart is also moving, this gives us the need to compensate for this issue in the image processing algorithm, for which we needed to simulate the Heart.

This makes me curious so as to know when given a problem, what is your methodology for approaching it?

Given any topic or problem, I always try to ask myself the following questions- what is the problem, why is it important to solve it, what are all the parameters that needs to be defined for its solution. Basically, we need to define all the inputs for the problem and try to come out with the optimal solution.

Where is the biomedical industry headed right now, i.e. focus of the companies?

This is a broad topic. There are different aspects of medical care. One is Preventive medical care. The next one is Pain Management Medical care, where you try to manage the problem when you don’t have the solution. Then, there’s this whole branch of field on DNA and specific protein markers-based treatment. Since, there are different sets of requirements in the medical field, each field is evolving on its own.

I would say that the game changer would be Personalised medicine. Every human being is different, although we have similar anatomy, our genes are different. If you take cancer as an example, one medicine will not work on everyone. There is a need for finding a specialised medicine for every patient. So, I think the future of Medicine is that way.

What domain skills would you expect a Biomedical Engineer Graduate to have, in order to work in the Instrumentation field?

Most of the research in the last couple of years, has been on Machine Learning, i.e. Artificial Intelligence. What the industry lacks right now is, Artificial Intelligence experts with background knowledge on Biomedical or any field you take for that matter. So, the educational system should impart skills in Machine Learning apart from the domain knowledge which each one specialises. The combination of these two would be the best.

My last question actually overlaps with part of your previous answer. Yet is there anything which you would like to add on to what steps you think the Universities can take in order to bridge the gap between Industries and Academic Institutes, given that nowadays, Unemployability is a major problem?

Yeah, given a chance that I could go back to University for education, I would want my courses to be flexible, where I can choose electives from multiple domains, where one of the courses I choose would be Machine Learning or Neural Networks based courses, apart from my regular courses. This would give me a distinct set of skill set on my resume which would have given me an advantage for my job in the interview.

Thank you again sir, it was a great pleasure to have you share your experiences.