You started your entrepreneurial journey in the early 2000s and co-founded Deeptek AI in 2017. Tell us about your journey.
My entrepreneurial journey began with a successful teleradiology company. While it was thriving, I recognised a barrier to further growth – limitations in traditional workflows and technology. This experience underscored the critical role of a strong technological foundation for scalability. It was during this time that I met Ajit Patil, who shared a similar vision of leveraging Artificial Intelligence (AI) in medical imaging. Our shared passion for bringing AI into radiology operations became the driving force behind DeepTek.ai’s inception in 2017.
What demand did you see in the medical industry to co-found radio imaging using AI? How is deeptek.ai revolutionising radiology imaging?
The global healthcare sector faces a severe shortage of radiologists. This is particularly concerning in India, with only 20,000+ radiologists serving a population of 1.3 billion. This scarcity leads to overworked radiologists, burnout and delays in reporting, impacting patient care.
DeepTek.ai tackles the global radiologist gap with a powerful AI deployment platform called Augmento. This US FDA-cleared radiology AI platform acts like a virtual assistant for radiologists, streamlining workflows and boosting efficiency. Radiologists can deploy any third-party AI model onto the platform, allowing them to use a single interface for various AI applications. Augmento prioritises urgent cases, meticulously analyses scans for abnormalities, and even generates reports with a click. This frees up radiologists’ time for complex cases and speeds up diagnoses.
For chest X-rays specifically, DeepTek.ai offers Augmento X-ray, another FDA-cleared AI tool. It sorts X-rays based on abnormality, allowing radiologists to focus on critical cases, boosting efficiency by up to 50 per cent.
Further, another tool named Genki tackles lung health challenges as a global lung health screening solution. This AI-powered tool analyses chest X-rays in just 15 to 20 seconds, identifying individuals with suspected lung issues like presumptive tuberculosis. This analysis is vital in areas with limited access to specialists, enabling early detection and faster treatment.
Take us through your journey of founding Deeptek.ai in terms of arranging investments and tell us when did you get your first profit.
We started with a strategic collaboration – my co-founder, a serial entrepreneur, leveraged his prior experience with NTT DATA to secure their early-stage investment. This funding fueled the development of our core AI technology. Following this initial tech development phase, we saw successful implementations in leading hospitals across India. Partnerships with established Japanese firms like Nobori and Doctor Net further broadened our global presence.
A defining moment came in 2019 with the Greater Chennai Corporation partnering with us for a TB screening project. This project marked the launch of our flagship chest X-ray AI solution and showcased the real-world impact of our technology. The success of this project propelled us towards the global market, and, in 2021, we secured Series A funding from Tata Capital Healthcare Fund II.
Tell us about any challenges you faced in your journey and how did you overcome those?
DeepTek.ai’s journey wasn’t without its hurdles. Here are a couple of challenges we faced and the ways we tackled them:
- Balancing innovation and adoption: Developing cutting-edge AI solutions is crucial, but ensuring their practical application in hospitals presented a challenge. We bridged this gap by focussing on user-friendly interfaces and functionalities that seamlessly integrated into existing radiology workflows. This ensured our solutions offered real value to radiologists without demanding a complete overhaul of their processes.
- Building trust in AI for medical diagnosis: There was a natural hesitation from some radiologists towards relying on AI for diagnoses. To overcome this, we prioritised transparency. We ensured our AI models were interpretable, allowing radiologists to understand the reasoning behind the AI’s analysis. Additionally, we emphasised AI as a collaborative tool, working alongside radiologists to improve efficiency and accuracy, not replace them.
- Limited connectivity hinders TB screening: Early on, our TB screening programme faced a challenge due to limited internet connectivity in remote areas. Chest X-ray scans were collected throughout the day, but results were only available when staff returned to a station with internet access. This delay meant confirmation by radiologists and initiation of follow-up steps, like collecting sputum samples, could take an additional day. This hampered our ability to quickly diagnose and isolate TB cases, potentially contributing to its spread.
Can Deeptek.ai integrate with the existing radiology machines, or does it require a set up of new machines to add AI?
DeepTek.ai shines by being vendor-neutral. Unlike some competitors that require specific machines, DeepTek’s Augmento platform integrates effortlessly with the existing radiology equipment. This eliminates the need for costly upgrades and minimises disruption to your workflow. Augmento’s compatibility with various devices and operating systems ensures a smooth adoption process.
Based on economic disparities in India, what steps do you take to make radiology incorporated with AI accessible to all sectors of people?
Addressing the economic disparities in India and ensuring that AI-enhanced radiology is accessible to all sectors of society is a multifaceted challenge. DeepTek.ai had worked closely to bridge this gap using the following ways:
- Government and NGO partnerships: We collaborate with government bodies and Non-Governmental Organisations (NGOs) to implement AI solutions in public healthcare systems. For example, our partnership with the Greater Chennai Corporation for TB screening showcased the potential of AI in public health initiatives.
- Mobile health units: To reach remote and rural areas, we deploy mobile health units equipped with portable X-ray devices and our AI-powered tools. These units can travel to distant locations, providing essential diagnostic services to populations that would otherwise have limited access to radiology services.
- Localised AI development: We focus on developing AI models tailored to the specific needs and conditions prevalent in different regions of India. By training our AI on diverse datasets that include images and cases from various parts of the country, we ensure that our solutions are accurate and effective for a wide range of pathologies and patient demographics.
- Continuous innovation: We are committed to continuous innovation, ensuring our AI solutions remain cutting-edge and accessible. By investing in research and development, we strive to improve the efficiency, accuracy and affordability of our AI tools, making those even more viable for widespread use across diverse economic settings. Through these steps, we aim to democratise access to advanced radiology services, ensuring that the benefits of AI in healthcare are felt across all sectors of society, regardless of economic disparities.
Reports suggest that exposure to radiology can cause cancer. Do you think using AI and other medical advancements can change the spectrum? If yes, how? If no, why not?
There are valid concerns about radiation exposure from medical imaging and its potential link to cancer. Patients often worry if radiation can increase their cancer risk. Indeed, the rise in the use of higher radiation-dose tests, such as CT and nuclear imaging, has sparked caution among experts. However, it is important to weigh the benefits and risks carefully.
CT scans and nuclear imaging have revolutionised medical diagnosis and treatment, reducing the need for exploratory surgeries and many other invasive procedures. When used appropriately, the benefits of these advanced imaging techniques far outweigh the associated cancer risks. On the other hand, AI has the potential to significantly mitigate the risks associated with radiation exposure in several ways:
- Enhancing accuracy: AI enhances the accuracy of medical imaging, which reduces the need for repeat scans. With more precise imaging interpretations, doctors can make more accurate diagnoses the first time, eliminating the need for follow-up imaging that would further expose patients to radiation.
- Early detection and preventive care: By aiding in the early detection of diseases, AI can reduce the need for more extensive imaging and treatments later on. Early and accurate diagnosis often requires less intensive imaging, thereby lowering overall radiation exposure over a patient’s lifetime.
AI is set to revolutionise the healthcare system in the coming years, with several key advancements anticipated as:
- Personalised care: AI will enable highly personalised treatment plans by analysing individual patient data, including genetics, lifestyle and medical history, leading to more effective and tailored healthcare solutions.
- Disease prediction: AI will excel in predicting disease onset by identifying patterns in vast amounts of patient data, facilitating early intervention and proactive care.
- Risk assessment: AI will refine the assessment of patient risks and vulnerabilities, providing personalised risk profiles and preventive strategies.
- Expanded telehealth services: AI will boost telemedicine capabilities, offering virtual consultations with AI-driven diagnostic tools, improving access to healthcare, especially in remote areas.
- Continuous monitoring: AI-powered wearables and devices will enable real-time health monitoring, alerting providers to any significant changes, and facilitating timely interventions.
dorina@thefoundermedia.in
