Healthcare startup DeepTek records over $1M in revenue within a year of starting up
Pune-based startup DeepTek focusses on radiology services by effectively leveraging AI to develop a platform that increases the productivity and quality of image analysis.
The prevention and cure of any disease depend largely on the correct diagnosis. And, radiology, a branch of medical sciences helps significantly in accurately diagnosing the disease.
Currently, a Pune-based healthcare startup DeepTek is leveraging new-age technology such as artificial intelligence (AI) to improve the quality and productivity of radiology services, and in the process, expand the reach of healthcare services to distant locations.
Founded in 2018 by Amit Kharat, Ajit Patil, and Aniruddha Pant, DeepTek is developing cutting edge AI tools for medical imaging.
“Our AI-based solutions – screen, triage, and diagnose thousands of radiology scans every day. AI is revolutionising healthcare by pushing the frontiers of technology,” says Amit Kharat, one of the Co-founders.
According to him, there are an estimated four billion people who do not have access to quality medical imaging expertise. Although teleradiology has managed to do the function of remotely diagnosing the radiology images, it faces challenges of quality, staffing, and scalability.
Teleradiology helps in transmitting a patient’s images such as x-rays, CT, and MRI scans, from one location to another to share studies with radiologists.
How it started
The individual expertise of the three founders, who are in their 40s, played a key role in starting up DeepTek.
With a DNB and PhD in Radiology, Amit Kharat has been engaged in the space for the last 17 years. His fellow Co-founder, Ajit Patil, has a Master’s degree from IIT-Kharagpur in Industrial Engineering and Operations Research, and has also co-founded a Japan-focused software services firm – Vertex Software.
With a PhD from the University of California at Berkeley, Aniruddha Pant is a data science expert and has considerable experience in leveraging deep learning for medical image processing.
Ajit and Aniruddha, who are alumni of the College of Engineering, Pune, knew each other from before, while Amit and Ajit met when the latter was evaluating teleradiology companies for his new venture.
“The three of us have successfully run multiple businesses, and we complement each other’s strength perfectly,” says Amit.
The AI edge for radiology
The founders of DeepTek, while looking at various challenges faced around teleradiology, realised that AI has the potential to overcome these problems.
Recent trends in technology such as availability of massive computing power and large annotated datasets have made it possible to build AI models that can assist and augment human radiologists and imaging experts.
“We realised that AI is no longer just an attractive differentiator, but it has become an indispensable technology for the entire medical imaging industry. That is when we decided to start DeepTek,” says Amit.
DeepTek offers a cloud-based picture archiving and communication system (PACS) platform, with built-in AI tools to triage and diagnose various pathologies through X-Rays, MRI, and CT scans.
According to the co-founders, “The platform gives customers the freedom to control the studies and scans they want to be reported on the platform. The flexible plans allow the imaging centres and hospitals to smoothly manage the workflow with a pay per use cloud strategy. This allows scalability with no extra costs or overheads.”
The AI solutions deployed by DeepTek helps in performing an "instant study triage" and also ‘flag' studies which AI finds as pathology. These studies are presented to radiologists and imaging experts first on the dashboard for their expert opinion on a top priority basis. This results in a faster turnaround time for radiology reports with critical findings. The ‘triage’ benefits are also instantly available to hospitals and imaging centres, who also have access to the triage dashboard.
“Our AI models diagnose over 15 medical conditions and produce auto-generated reports. The software provides imaging experts with the option to approve these reports or make modifications to them. These modifications are again given as input to the AI models, forming a feedback loop that allows our models to continuously learn and improve,” explains Amit.
For instance, for digital chest radiograph, the startup assists in flagging suspected studies of TB, normal blood vessels, lung nodules, pleural effusion, consolidation, lines, tubes (ICU Radiology Workflow Optimisation Solutions), and many such conditions.
DeepTek is also part of GE Healthcare’s first startup cohort programme on its Edison platform.
Initially, the healthcare startup faced challenges as conceptualising a strong AI-enabled platform and bringing it to the action required seamless contributions from data science, technology, and radiology domains.
The challenge primarily stems from the flow of data and creating a model that can provide accurate analysis in a shorter duration.
“This required cross-talking to happen seamlessly across the domains rather than within domains. There were initial hiccups, however eventually, things fell in place,” adds Amit.
DeepTek co-founders claim their repeated trials demonstrated that the platform improves productivity by 86 percent and reduces the average reporting time by 46 percent. “Our thoughtfully designed user experience ensures that the radiologist can process studies with minimum effort,” says Amit.
DeepTek’s pay per use platform allows hospitals to deploy instantly, scale rapidly, use AI and deep learning tools to make reporting standardised, and thereby, get consistent qualitative reporting.
Currently, DeepTek’s team has around 45 employees which include radiologists, data scientists, and software developers.
At present, DeepTek’s annualised revenue rate (ARR) is over a million dollars. According to Amit, the platform is being used across 50 hospitals and imaging centres globally. For example, Greater Chennai Corporation’s TB Free Chennai Initiative is using the startup’s technology platform.
The positive customer traction is a big vote of confidence for DeepTek. Amit says, “We are one of the few AI companies in this domain that have achieved commercial adoption of their products and services. We have loyal customers and are generating revenues.”
Given the diverse skill sets of the co-founders, DeepTek can understand the challenges and opportunities on all fronts, be it business, radiology, or technology. This has given the startup an edge in the marketplace which has several global imaging companies such as Siemens, GE, and other startups working in the area.
DeepTek has raised a seed funding from NTT Data, a Japan-headquartered IT services giant and a strategic investor in the startup. It has also raised funding from other strategic and institutional financial investors.
Moving forward, Amit says, “Building a world-class team, getting statutory clearances, adding to the portfolio of pathologies our platform covers, continually improving our user experience, serving more customers, and ultimately saving more lives will be our focus moving forward.”
(Edited by Suman Singh)