News

Updates and analysis on technology, impact investing, and our venture companies.

Next Page Navigation Link
Previous Page Navigation Link

Much of current medical standard procedure is trial and error. You’re feeling ill, you relay your symptoms to the doctor, and your doctor then prescribes you a treatment—the same exact treatment that’s prescribed to anyone with your symptoms. And if it doesn’t work, you’re prescribed another one. And another. But what if you could get it right the first time?

Instead of a one-size-fits-all method, precision (or personalized) medicine is tailored to fit each individual. Using this approach, doctors take into account not only current symptoms but also patient history, environment, genetics, and lifestyle in order to treat and prevent the disease. The generalized care system is moving towards “more effective, targeted drugs; less uncertainty, more accuracy—and, ultimately, better care.

Although precision medicine is most often associated with genomics, mobile technology and health applications can also play an enormous role in individualized care. As consumer digital technologies become more sophisticated, a vast amount of data is becoming available that creates a more accurate picture of each patient. “We’re heading toward being able to do your own medical selfie,” Eric Topol, a leading cardiologist and professor of genomics, says in an interview with MIT Technology Review. And with the right tools, those “medical selfies” will be extremely useful for precision medicine solutions.

But precision medicine isn’t without its problems. First of all, it’s expensive and time-consuming. That has made it slow to deliver on its promise of transformative outcomes. Consider ivacaftor: according to one writer at the Scientific American, although the drug eased symptoms in 5% of cystic fibrosis patients, it also took decades to develop and costs $30,000 per year per patient. Worse still, it functions similarly to three other, cheaper medications already on the market (including high-dose ibuprofen).

Personalized treatments require enormous amounts of data, and that presents a challenge. Collecting the data can be burdensome for patients, and parsing the data is tedious and time-consuming for doctors, even with algorithms to help automate the process. However, we’re approaching major advancements in both mobile healthcare technology and artificial intelligence, and that will soon help deliver precision medicine treatments at unprecedented scale.

At Aspire, we’re working on advancements in multiple fields of healthcare using a very unique approach to artificial intelligence. The problem, we believe, with most AI approaches to personalized medicine is that they still use a one-size-fits-all algorithm. We think it’s a fool’s errand to employ the same static algorithm for every patient and expect personalized results when everyone’s biochemistry and behavioral patterns are dynamic and unique. Our adaptive artificial intelligence platform, A²I, assembles algorithm components on the fly to create a fully personalized model. This approach means that not only is the data analysis fully tailored to each individual patient, it also means that the algorithm is always self-optimizing and always adapting to new variables, new data sets, and new inputs.

One of our ventures, Tempo Health, has achieved some exciting results in their work with diabetes management using A²I that we believe could point a way forward for scaling other precision medicine solutions. Tempo Health recently partnered with a clinic in the Netherlands called Diabeter on a joint study, and the findings of that study have been very promising.

Diabeter is one of the largest diabetes specialist centers in Europe focused on providing individualized care and 24/7 support to help children and young adults with type 1 diabetes better manage their blood glucose. The doctors at Diabeter remotely monitor data from each patient streamed from CGMs, insulin pumps, and other devices on a regular basis, and in some cases doctors even make adjustments remotely to patients’ insulin pump dosages based on the data. This high-touch approach has achieved truly impressive results, with a hospitalization rate of only 3%, compared to the Netherlands’ average 23% hospitalization rate for diabetes patients. But offering around-the-clock support and individualized treatment plans takes a team of medical specialists and a lot of resources, and that makes their model incredibly difficult to scale.

That’s where Tempo Health comes in. In their study with Diabeter they used our unique adaptive artificial intelligence platform, A²I, to see if we could automate parts of their data-tracking and analysis model without sacrificing outcomes. Not only was A²I able to match the doctors’ results, A²I actually made improvements--with a 20% increase in patient time spent in the safe blood sugar range and 9% fewer hypoglycemic (dangerously low blood sugar) events.

Better yet, these results are scalable. With the introduction of A²I in a wearable glucose management solution, another project that Tempo is working on, we could dramatically reduce the hospitalization rate and save billions in diabetes-related healthcare costs every year.

The observational study is just the beginning. Next, Tempo Health will be developing a closed-loop artificial pancreas system called Rhythm that will take advantage of our sophisticated A²I platform. And that’s just the tip of the iceberg. With a dynamic, adaptive AI solution, we can use our same A²I platform to scale other precision medicine solutions in a vast array of other fields in healthcare--from medical imaging to personalized physical therapy assistance.

Precision medicine is the next step in the transformation of healthcare, and A²I, we believe, is the next step in transforming precision medicine.