Sept 30, 2022
How did all this begin? An Interview with CEO Kovey Kovalan
Back in grad school, I worked on a NASA contract aimed at visualizing data. To help with the work, my professor purchased a $30,000 research computer. I thought the computer was too slow so I started digging into the technical architecture of the machine. As I explored the machine, I realized that it was similar to a computer gaming system. I bought a $200 Video Gaming Graphics card off-the-shelf in the hope that I could tinker with the Graphics card architecture to make something faster and better. Within a few days, my $200 graphics card was outperforming my $30,000 research computer. Video Gaming Graphics cards use thousands of processors which we now call a Graphics Processing Unit (GPU).
After graduation, I took a few jobs in AI and videogaming. I thought I had hit the jackpot when I was offered the chance to work on the visualization team for the #1 medical imaging enterprise system in the US. I brought the graphics processing units I had tinkered with to the decision makers at this company. I thought these off-the-shelf GPUs had enormous potential for relatively low cost. But my bosses weren’t interested. Eventually, I left because the organization remained committed to old technologies and they didn’t share the passion I felt for the growing computing power of the off-the-shelf GPUs. And I couldn’t shake the feeling that if we didn’t embrace the power of off-the-shelf GPUs, someone else would come along and totally disrupt the industry with the power to enable advanced visualization in medical imaging.
After I left that #1 ranked company, I kept tinkering with the GPUs. I decided to go to RSNA to show off what I thought the GPUs could do. I didn’t want to carry my GPU server with me so I built a web-based application that could access the server remotely on any device with internet access. I could then showcase these large imaging data sets without lugging around my GPU server.
At RSNA, I met the Medical Director of Sharp Children’s Hospital. He was shocked that I could display these large medical imaging data sets on my phone. At the time, he couldn’t move even one data set between floors at the hospital.
The Medical Director hired me to create a pilot program for Sharp Children’s Hospital. We built a whole GPU-based system for medical imaging remote access. The servers were on-site at the hospital but they could be accessed anywhere through the web application. This was the first remote access system of its kind. But once we deployed, the speed and latency weren’t as good as I knew they could be.
I investigated and realized that because the on-site servers were connected to the hospital’s Local Area Network (LAN), there was too much daily traffic clogging the network’s bandwidth.
I believed the answer was to host the medical imaging system off-site at a secure data center. The hospital decision-makers were very concerned about privacy and medical record security. I took them to the secure data center in San Diego where Pfizer kept all their data. When the decision makers saw all the security protocols in place at the center where Pfizer hosted their data, they became convinced that the medical records would be safe—or even safer—at a secure data center.
Once we moved the server to the off-site secure data center, we created new applications that could communicate with the on-site modalities but could also be accessed remotely in real time during emergency situations. Once we updated the applications and deployed the off-site servers, the system achieved record-breaking speed and zero-latency.
That was the AHA moment – that was when I saw that this was a new approach to medical imaging. An off-site, hosted solution using GPUs could really work with all on-site modalities and legacy systems and also deliver speed, zero-latency, remote accessibility from any internet-connected device, HIPAA-compliant security, and patient privacy with no on-site hardware. To describe this new concept, I used the term “web portal.” No one was using the term cloud back then.
I started to get excited about this web portal concept I had implemented for Sharp. But when I told other people in the medical imaging community about it, they laughed at me because they were so convinced that patient privacy and security could only be protected on-site. But I knew security was better with my system. And I saw that I had to convince hospitals to move away from on-site systems and into a system like mine.
So now we get to May 2009. I had just self-funded the Sharp pilot that proved the concept beyond what I’d imagined. But I was broke and I knew I had to file the patents on the system. So my wife gave me her credit card and said: go file the patents.
We were struggling to put food on the table. There was a big recession and no one was funding new projects. But my wife believed in me when no one else did. I filed the patents on my birthday—May 28, 2009. A month later, I landed my first investor—in the middle of this recession. And subsequently, I won a National Science Foundation grant to research using GPUs for AI in medical imaging.
For the next few years, we developed the technology that makes today’s product possible.
First, we figured out that in order to work with the RIS-PACS vendors, we had to position our zero footprint viewer to integrate with their existing system. The viewer needed to satisfy the full functionality of a radiology workstation but run from our cloud with remote visualization capabilities. This was a ground-up approach because raw images could now be transmitted to the end user without requiring any processing capability from the end user device.
We went back to RSNA in 2011. There, we were recognized as the first zero footprint viewer in the Daily Conference Bulletin. Then, Frost & Sullivan got wind of our effort and they awarded us a technology innovation award.
We soon recognized that in order for this technology to be fully utilized by practices with legacy RIS-PACS, it would need to be integrated into the existing workflow for all the stakeholders at an imaging center. And that’s what we created—a distributed cloud-based system that could completely replace an on-site system if desired.
In 2014, we were granted our first patent for a broad scope generic system that includes the entire conceptual framework for what had been implemented and what was possible to implement.
Once our patent was granted, the rest of the industry started using similar terminology: “Zero footprint! Cloud-based! Interoperability! AI-based!”
At first, I wondered—had they figured out another way? One by one, I dug into their architecture. One by one, I saw that they were using the same words but none of their platforms came close to what we offer.
Plenty of companies tried to duplicate what they thought our platform provided. But none of them could figure out how to deliver what we deliver.
In 2016, RSNA invited us to showcase our image-sharing interoperability technology at their image-sharing interoperability booth at no cost.
In 2017, we landed our first complete replacement of a medical imaging system for an imaging center in Brooklyn. We had displaced one of the top enterprise RIS-PACS. It was a market-winning milestone for us.
We added a few more imaging centers along the way. In 2020, we landed a distributor agreement that accelerated our growth to an additional 40 imaging centers. Our platform now serves 22,000+ referring physicians, 275+ radiologists, and 1 million+ patients.