放射線医学は撮像のデジタル化によってペーパーレスとフィルムレスを達成し、撮像のデジタル化は場所に依存しない放射線業務を可能とした。デジタル化を通じて放射線科医師の業務に大きな変革をもたらしたと言える。しかしマンパワーに依存する読影業務には作業量に限界がある。人々の放射線医学へのアクセスをより簡単にする為に、AIを用いた解決策を模索するスタートアップ企業がしのぎを削っている。その中でも特に注目を集めているのがDeepTek社だ。共同創設者で放射線医師出身のDr. Amitが率いる同社は、AIを用いた放射線科業務の効率化によって読影業務のみならず放射線業務全てに関わる作業の効率化を図り、より多くの人々に放射線医学の恩恵をもたらそうとしている。今回、Dr. Armit氏が取り組む放射線医学の「民主化」や、その道のりと今後の展望についてインタビューを行った。
The “journey” from a radiologist to a technology enthusiast.
——What is your motivation to become a doctor?
Dr. Amit My entire family has doctors-orthopedic, ophthalmologists and dentists. I was fascinated by my elder brother and my sister to become a doctor both of whom are remarkable at what they do. The choice of sub specialization came later in my career. When I was interning in the radiology department, I was really amazed by the technological innovations available in the hands of experts in the field of radiology. This fuel my imagination.
——Which university did you graduate from?
Dr. Amit I belong to Pune city, which is also known as the Oxford of the East. I am an alumnus of Fergusson College, Pune. I pursued the foundation medical course from B.J Medical College, Pune which is a part of Sassoon Hospital Pune (founded in 1878). Subsequently I pursued my post-graduation from Rural Medical College, (then a part of Pune University). Following this I did my DNB and PhD in Musculoskeletal Imaging from D Y Patil University, Pune.
——What is your reason for aiming to become a radiological technologist?
Dr. Amit Technology is fascinating and evolving. I am inspired by the visionaries like the American Physicist Michio Kaku, the late American Astronomer Carl Sagan, Indian Physicist C V Raman and mathematician S Ramanujan. They inspired us to give your very best to achieve your goals and mission. Radiological equipment are the epitome of perfection and marvel achieved through cutting edge technological mastery which test the limits of physics and medical science. Since early days of my career, I was fascinated with Telehealth and Telemedicine and its applications in radiology as it has the power to truly democratize radiology. This propelled me to work towards adopting and validating the latest tools like cloud and data mining to ensure high quality report generation for hospitals and imaging clinics I was serving as an expert. Being a medical domain expert, I had limited technological knowledge and hence learnt these new tools through consults with experts in these fields and then adopt these to transform healthcare delivery in the Indian Subcontinent. The learnings from these actually helped me write research papers in peer reviewed journals. This started my journey allowing me to learn new technological solutions fueling radiology like DICOM standards, cloud, data mining, big data and AI in radiology. This is how my journey progressed from a radiologist to a technology enthusiast.
Driving AI in almost all the aspects of radiological practice.
——What is your reason for a company start-up of AI Healthcare in radiological diagnosis?
Dr. Amit As I was evolving my teleradiology practice I was successful in creating a boutique radiology practice in western India catering to a particular modality and group of customers. However, we saw that the demand was higher than what we could handle. However, scaling teleradiology to cater to the increasing demands was not easy. Teleradiology is highly stressful and the pressure is akin to the pressure the air traffic control team might be facing on a daily basis. The stress is visible at operations, management and even radiology reporting aspects. This resulted in burnouts amongst experts. I saw that this story was repeatable across similar other organizations which are trying to scale their practices. Through rigorous quality control organizations were managing to keep the checks and balances. However, to scale beyond this to manage demands required a completely new innovation. AI was the missing piece here. AI not only for medical image analysis, but also for workflow automation and smart report generation and report distribution. It was a coincidental meeting with Ajit Patil and his interest in this space and his experience in the field of technology development and research made me confident and pleasantly surprised that together along with a strong leadership team this goal is achievable. That led us at DeepTekto develop Augmento – the industry first AI PACS. Using this solution imaging clinics and hospitals can now obtain a productivity gain of 2x to 5x and at the same time ensure high quality reports with minimal errors. The solution is assistive technology akin to building an auto pilot for the aircraft, we are basically building a virtual assistant for almost all aspects of radiological practice. This journey actually got conceptualized as DeepTekand focus moving ahead is to help driving AI in almost all the aspects of radiological practice.
——Please describe the success story.
Dr. Amit DeepTekhas been successfully deployed its Chest AI solutions and AI PACS Ubiquitous Radiology platform “Augmento” in the clinical workflows across 400+ imaging clinics and hospitals who are paying customers. However, our key deployments were establishing our work in the infectious disease screening space using AI in chest x-rays. It was a crucial deployment in a large and populous state and we could execute it successfully considering the odds of limited infrastructure available on site and even connectivity issues like stable mobile networks and internet connectivity. Now using the tool, we are in the third year of deployment and we have already screened 150,000 patients using DeepTekChest AI Solution.
——The technology developed by DeepTek is attracting attention from all over the world. Please tell us why it has been successful.
Dr. Amit There are many companies developing AI and each AI has huge potential. A huge respect to all of them. However, no one is addressing the problem of deploying AI in Radiology workflows. DeepTekis an AI first radiology Picture Archival and Communication System (PACS) and also a vendor neutral archive (VNA) capable of understanding the complexity and innovate on solutions on ease of deployment of AI in radiology workflow at a click of a button. Today radiologists, imaging clinics and hospitals are adding in one AI at a time in the radiology workflow. This way the journey will take ages. DeepTekaddresses this issue by creating an AI Deployment Platform which is a thin client, seamless and ubiquitous solution which can add the key aspect of integrating 100+ AI solutions in the radiology workflows. Hospitals can plug and play AI and also easily orchestrate AI in the workflows and this differentiates DeepTek. Being vendor neutral and cloud agonostic we can deploy the solution on any cloud and in any geography. The solution can also be deployed on-premise if the hospital guidelines makes it a must have.