AI, Bloodmobiles, and Confidential Computing: The New ABCs of Precision Medicine
Written by Sheryl Tullis
Curing cancer has long been the brass ring of the medical field. With emerging data security technology, that outcome may be on the horizon at last.
Lisa Thee, Launch Consulting’s Data for Good Sector Lead, is on a mission to improve health and wellbeing for all through digital transformation. Accelerating clinical usage of AI to complement the expertise of doctors and utilizing precision medicine, Lisa believes, will help increase accuracy in diagnosis, promote recovery, and improve outcomes for all patients.
The Rise of Precision Medicine
For Lisa, who mainly works with tech sector leaders to mitigate social evils like hate speech, terrorism, and child trafficking on their digital platforms, healthcare transformation is a bit of an outlier. “I got energized by seeing how much progress was being made in precision medicine for cancer treatment,” she says. “Your genes, behaviors, and environment are all factors that affect your health. By measuring these factors and acting on them, interventions can be tailored, rather than using the same approach for everyone. And AI is accelerating precision medicine.”
“The cure for cancer won't come in a pill; it will be early detection.” – Ravi Patel, PHD Global Technical Product Manager, Kheiron Medical Technologies
The idea? If data stewards like research hospitals could securely provide data to algorithm owners like startups, academic researchers, and medical device manufacturers, then model validation and training on more diverse datasets could accelerate—without ever having to move the data. Since data is most vulnerable to attack when in motion, this solution would solve the security issues of using this valuable, regulated, well-guarded data. Additionally, with machines performing the tedious, repetitive work of model validation, doctors would be freed up to focus on seeing patients.
“In order to accomplish this goal,” explains Lisa, “we needed to create custom solutions for confidential computing, with a frictionless user experience that preserves access for people who need more diverse data, while also maintaining the IP for healthcare systems.” An important part of the development process also included addressing how to reduce bias in models. The team aimed to ensure that the models perform as well “on a middle-aged man in India as they do for an elderly woman in Indiana.”
The Right Place, The Right Time
For healthcare organizations that have a strong desire to use their data for something innovative, but for which the risks have outweighed the potential benefits, confidential computing is the key to creating an environment where their ideas can thrive. As Launch Architect Charles Cozad puts it, “When you have a system that no one can see into, and you can bring together data that you’ve never been able to have side by side, that’s where the magic happens.”
This confidential computing project isn’t Launch’s first experience with breakthroughs in the healthcare industry. A few years ago, an innovative biotech company sought to provide precision oncology treatments for prostate cancer patients. The company had created a treatment that stimulates patients’ immune system to fight cancer cells, but it required a rigorous process: the patient would have their blood drawn at a doctor’s office; then, the blood would be shipped via “bloodmobiles” to a filtering center to isolate white blood cells, which would then ship off to another location to incubate with the drug for a couple of days. Finally, the revved-up cells would be returned back to the clinic for reinfusion into the patient.
“When you have a system that no one can see into, and you can bring together data that you’ve never been able to have side by side, that’s where the magic happens.”
Keeping the blood safe, at temperature, and tracked during its jaunt around the country was no mean feat. Losing the blood, or delivering the wrong blood to the wrong people, would be a matter of life and death. Launch helped that company pioneer the first FDA-approved process for the logistics and timing of collecting, treating, and delivering individual patients’ blood samples door to door.
On their faces, the bloodmobile delivery model and the confidential computing model may not seem related—but they’re more similar than they appear. Whether samples are locked down in a bloodmobile or locked down in a zero-trust computing environment, precision healthcare is about security, regulatory compliance, individualization, and speed. It’s about the right information, in the right place, at the right time, to create the right outcome for the right person.
Despite growing adoption of tech innovation and tools around the industry, most healthcare still happens in the physical world. As AI and confidential computing gain steam, it will be vital to translate those models to analog treatment and recovery. Together, these capabilities will allow healthcare organizations to combine the power of patient data with the power of preventative, proactive, and personalized care.
No one believes this more strongly than Lisa Thee, who is emphatic about using tech as a tool to connect more wholly as people. And when it comes to using confidential computing, specifically, as a tool to cure cancer through early detection: “I have never felt prouder of a team coming together to protect our human legacy.”
Lisa thee Data for Good Sector Lead Launch Consulting