The Power of Quality Data

Don’t Spread Your Research Thin

CorVista’s early work was processing physiological signals—electrical, hemodynamic, and imaging—with a team of data scientists who had no prior experience in medical devices. They collaborated with a doctor, gathered signals, and, on a modest $2 million budget over two or three years, demonstrated that their technology could potentially detect abnormalities in the body. 

Achieving this in such a short time and with limited resources was no small feat. Don credits this success to the team’s commitment to remain focused. While the company holds patents in cardiology, neurology, and oncology, they chose to focus on cardiology, because, in medtech, resources and time are limited. As Don puts simply,” The key to a small company is to stay focused.” 

So, how do you decide how to narrow down your focus? For CorVista, their research pointed the way. The team identified coronary artery disease (CAD) as not only one of the most critical global health issues but also an area where their technology could make a significant impact. Despite decades of treatment options, CAD remains a leading cause of death globally. The efforts in reducing mortality rates have unfortunately made limited progress. This lack of impact is what drew CorVista’s attention to CAD. Don explains, “If you take the energy, the three-dimensional electrical energy produced by a heart, you can detect signals that are being released from the body; we could interpret those signals and see a difference between a disease state versus a healthy state.” 

This focused approach also improved resource efficiency. If Don hadn’t chosen to concentrate on a single indication, the team wouldn’t have been able to demonstrate hard evidence of their technology’s capabilities with just a $2 million budget. 

Moreover, a focused strategy helps maintain regulatory and quality standards. At CorVista, implementing a quality management system early on—specifically tailored around their target indications—ensured consistency and compliance throughout product development. Don emphasizes, “From the time I joined CorVista, everything was done through a quality management system, almost as early as employee number two or three.” 

In short, by zeroing in on cardiology and building a solid foundation early on, Don was able to develop a powerful, targeted solution with limited resources. Thanks to their concentrated efforts, CorVista filed 180 patents, with 81 already granted—a substantial base of intellectual property built from the early days of AI and machine learning in an emerging field.

Commit to Clean Data for Clearer Diagnostics

“The key is that in data science, poor data is like garbage in, garbage out. But if you put in really great data that you collected, you know exactly where it came from and that it was clean and clear, then you can actually obtain or decipher information about that data very well,” Don shares. This focus on data quality has been central to CorVista’s mission of creating a technology that accurately diagnoses CAD and PH. 

You don’t have to be a data scientist to appreciate the value of high-quality data. Without it, you risk compromising the accuracy and credibility of your findings. To avoid this, CorVista made a decision early on—they managed their trials internally instead of outsourcing to CROs. Don explains, “We did so internally because the data was so important; we had to make sure that data was pristine.” Keeping this function in-house gave CorVista full control over the data collection process, and the integrity and precision they needed. This also allowed CorVista to build direct relationships with end users—physicians and clinicians who would eventually rely on their diagnostic system. Working closely with their future users, CorVista could gather valuable feedback and insights while strengthening trust and understanding within their target market. 

The sheer volume and precision of the data collected were also critical. Through FDA-sanctioned clinical trials, CorVista gathered information from over 11,000 patients, building a robust dataset to train and refine their algorithms. This extensive database allowed them to differentiate between healthy and diseased states with impressive accuracy. “We can detect significant coronary disease...and the lack of significant blockage with a 99% negative predictive value,” Don states. This high level of accuracy is only achievable because of the well-structured, high-quality data they collected. 

Today, the CorVista System delivers immediate, actionable results. In a doctor’s office, the test takes only three minutes, with no invasive procedures or radiation involved. Don explains that their system collects over 10 million data points per patient and processes these through cloud computing to generate a report that helps physicians diagnose conditions like CAD and PH. The simplicity and speed of this workflow make advanced diagnostics accessible even at the primary care level. 

In short, when it comes to your approach to clinical studies, make every effort to maintain control over the process to guarantee the quality of data. Managing trials internally ensures not only clean, reliable data but also fosters direct engagement with end users. By being on the ground with it, you gather valuable insights, build trust, and refine your technology based on real-world feedback. With quality data and user engagement in place, everything else—whether solving a problem or training an algorithm—becomes much easier.