Disease and pain are complex phenomena, and many people have problems that are difficult to treat. Innovative therapeutics that improve conditions can be equally complex, and it’s often challenging to calibrate their application to a patient’s individual needs and unique response to treatment. Personalized, precise medicine seeks to do that. One key to its success is artificial intelligence (AI) — the ability to continuously “learn” from data and outcomes and, in the case of medicine, tailor and fine-tune therapies for specific patients, so they can get relief and get well.
Take the work we’re doing around gastroparesis, or paralysis of the stomach. The condition is characterized by an inability to process and move food out of the stomach into the intestinal tract in a reasonable amount of time. This problem can cause bacterial issues, unrelenting nausea and vomiting, epigastric pain, reflux disease, indigestion symptoms, and heartburn — as well as anxiety and depression, which can exacerbate the symptoms. The lifelong disorder is more common in women than in men, usually has an onset in young females in their teens and early 20s, and can worsen progressively over the course of the patient’s life.
It’s estimated that from 2.5 million to 4 million people in the United States have gastroparesis. The condition is significantly under-diagnosed, and there are few good treatment options. One treatment that seems to relieve the nausea and vomiting is a gastric nerve stimulator that targets the vagus nerve in the stomach (this is one of the nerves that connects the brain to the rest of the body).
The trick is to measure the delicate relationship between the amount of stimulus applied to the nerve and how the patient feels in response to that “dosage.” My lab developed a way to gauge this through the skin surface — we were the first in the world to identify a relationship between stimulation, nerve response, and a change in patient perception of symptoms.
We then created AI algorithms to both predict how the nerve will respond to any stimulus and control individual pathways within a nerve. The algorithms are designed to improve the way any nerve stimulator works. In this instance, we’re applying it to gastroparesis by stimulating vagal nerve responses and verifying that we’ve gotten the correct responses on a patient-to-patient basis to better predict patient response to the therapy.
This is a learning algorithm, which means we use feedback from the patients — symptom scores — to adjust the stimulus and treatment in real time. Essentially, we’re using data science to help understand the condition from a crowdsourcing perspective. If individual patients consent to share their data, we can derive insights to unlock the underlying mechanisms at work to better inform the design of future and better medical devices for these patients.
My vision and my dream for this whole field is that any neurostimulation solution in the future will be as predictable and easy to take as a drug therapy and that it will have controllable, predictable side effects. There are numerous applications for these nerve-stimulator medical devices — for conditions like chronic pain; visceral pain; irritable bowel syndrome; epilepsy; depression; incontinence; and autoimmune disorders, such as rheumatoid arthritis.
Our research is part of the National Institutes of Health (NIH) Common Fund SPARC program. SPARC, which stands for Stimulating Peripheral Activity to Relieve Conditions, aims to map out what different nerve circuits do to internal organs — like the heart, lungs, stomach, intestinal tract, colon, and bladder — to speed the development of therapeutic neurostimulation devices.
Gastroparesis is a rare disorder, and there’s not a lot of money or effort being pumped into understanding or treating it. I want to change that because I feel very compassionate toward these patients. They deserve a better treatment option, and I believe we have only begun to fulfill the potential of our solution.
Matthew P. Ward
Assistant Professor, Weldon School of Biomedical Engineering
College of Engineering, Purdue University
Adjunct Assistant Professor of Medicine, Indiana University School of Medicine
Co-founder, Drug Free Therapeutix, LLC