Shahid Akhter, editor, ETHealthworld, spoke to Ajit Patil, Co-founder, DeepTek, to find out about the transformation in radiology through deep learning that has the power to enhance image analysis and reduce turnaround time.
What challenges and opportunities do you see in AI and radiology?
Radiology is an emerging industry. There is a huge potential to bring AI into the radiology workflow and transform the scenario. AI is powerful in image analytics, image processing, and radiology. It is largely about analysing images, processing images, and identifying pathologies based on what you see in the images. And hence, I think AI is well fitted in this space and can bring transformation.
The opportunity to bring in advanced image analytic techniques through AI into the radiology workflow has a huge potential. This has emerged over the last 10 years. AI technology came forth and it enabled to analyze images in a very cost-effective and efficient way. More than 100 global startups have been funded in this space only. And then radiology is a wide space. There are different types of modalities catering to different body organs. There are so many different types of pathologies. And these startups are independently working, trying to cover various areas within the radiology space.
A few years back, it was thought that AI could actually transform radiology and make it autonomous. And there was a lot of concern and fear. With the passage of time, a lot of clarity has emerged that AI will be only an assistive tool. AI will help radiologists improve productivity, improve quality, reduce turnaround time, reduce errors, reduce fatigue, and help radiologists have a better work-life balance. In that sense, now, radiology AI is evolving, and there are several several areas. There is an opportunity in emergency radiology where a patient through trauma, accident, or some other incident has come into the emergency room late at night and you want to quickly identify the challenges. In fact, an immediate medication could be delivered to save the life of the patient. There is public health screening also, wherein you have diseases like tuberculosis, which are widespread and you need to automatically identify on test X-ray imaging whether the person is infected or not. As a result, I believe that various opportunities are emerging as a result of AI technology to see the kind of transformation that exists in radiology workflow. This will bring much needed improvements, making radiology more accessible, making radiology more affordable, and making radiology more accurate.
What is the impact of Deep Learning Solutions on this industry?
It was in the year 2000 that deep learning solutions started having an impact on the industry. Deep learning has the amazing power to analyse images and identify objects or segment objects in those images. Radiology is all about analysing images and identifying lesions in those images. And it was an excellent match. It was Stanford University that brought in the early algorithms that analysed chest X-rays. And since then, a lot of evolution has happened. Now we have deep learning algorithms to analyse not only cross-sectional images like CT scans, MRIs, ultrasounds, PETs and others. And over that period, algorithms have evolved to become more mature and more robust. Also, regulatory approval mechanisms, which were missing earlier, have started to take shape and we now have seen some of the algorithms getting regulatory approvals as well.
What are the current deployments of DeepTek platforms?
One important deployment of technology is in Singapore. Singapore government through its National Healthcare AI strategy, are looking for an AI platform that brings AI to radiology workflow in a very responsible way, and we were evaluated with quite a few other competing legacy firms in this industry. We won the bid and our platform today is getting deployed in almost five public hospitals in Singapore. Our platform will become the de facto platform for AI in a very seamless in a very responsible way. And we are looking forward to this successful use case to many other countries from here. Not many radiology platforms are having clinical and commercial adoption. DeepTek has been working to enable this. We have been lucky to have strong support not only from customers but from industries ass well. So firms like Tata Capital Healthcare Fund, and quite a few other global firms have come together to support us through strategic equity investments. Besides, some of the large global IT firms from the large medical imaging device makers have been supporting us with the go to market strategy as well.
What are the market opportunities for DeepTek?
With this solid foundation of global customers, global partners, and a strong team of 100 plus people now, I believe we are amongst the prominent Indian players here. But there is a huge global market opportunity, and our vision is to see how we can really take the technology shared in global health, help radiology to become more accessible, more affordable, more accurate, and, in that sense, deliver solid improvement in patient care. We hope to be one of the leaders from India who can play a very important role in this market.