Artificial intelligence (AI) is becoming increasingly embedded in our everyday lives, often in ways we aren’t even aware of. From streaming services suggesting what to watch based on previous selections to GPS apps recommending the best routes to take factoring in current road conditions to ChatGPT compiling a list of potential examples that could be used to introduce this very blog.

In the simplest of terms, AI is the ability of a computer to think and learn so it can perform tasks typically done by people that require things like comprehension, analysis, problem solving, and decision-making. However, AI can process huge amounts of data at speeds humans can’t, making it an exciting new tool that could play a significant role in improving ALS treatment and care, starting with reducing time to diagnosis.

Currently, ALS is a diagnosis of exclusion, which means doctors have to rule out other conditions with similar symptoms before they settle on ALS. As a result, it can take a year or even more of sometimes redundant and painful diagnostic testing and even erroneous diagnoses and unnecessary treatments, such as spine surgeries, carpal tunnel releases, and immunotherapies, for someone with ALS to receive the correct diagnosis.

This diagnostic delay can result in loss of physical function, lack of access to disability benefits, and missed opportunities to participate in clinical trials.

“If we could have a solution that could accelerate this time to diagnosis, can get patients to that answer much, much, much faster. It could have long-term impact on the outcomes for ALS patients, their caregivers, and the broader community,” Nadejda Leavitt, Senior Director, AI for Healthcare & MedTech at IQVIA, told ALS Nexus conference attendees during a plenary session about “The Future of AI and ALS.” During this session, Leavitt shared the stage with Brenton Hill, J.D., head of operations and general counsel for the Coalition for Health AI, and moderator Dean Feener, chief information officer for the ALS Association.

Image

One promising way to do this is to “stand up a predictive AI algorithm.” A predictive AI algorithm, Leavitt explained, “would look at the medical history of many, many patients who are diagnosed with ALS and try to find what is unique about the footprint that they leave in EMR [electronic medical record] data. For example: What are some of the early symptoms? What are some of the misdiagnoses? What types of providers do they see? What do they mention to the clinician that ends up in the clinical notes?”

“An AI algorithm can mine all of this data and try to identify other patients who do not yet have an ALS diagnosis but might be undiagnosed,” Leavitt continued. “It would then present this information to the clinical care team, along with supporting information … and the recommended next steps.” 

Image

While such an algorithm or model is currently hypothetical, the data needed to create it already exists.

“Every time you go see a doctor, there's a lot of diagnosis codes or procedure codes, treatments that get inputted into the medical record,” Leavitt said. “The clinician would also write a lot of notes where they would mention things that you said in passing. This type of data is available to an individual health system. It also often gets aggregated into national research databases. So, everything that is already passively being collected … could be used for AI algorithm development.”

As a real-life example, Hill highlighted an AI algorithm developed by Dr. Nathan Staff at Mayo Clinic that has been trained on 45,000 EMGs (electromyographies) to help diagnose ALS and predict survival.

“I think ALS, historically, has been a disease [that’s] not well understood,” Hill said in a separate interview. “And so, by providing that type of insight to people, it really gives them a sense of, ‘Okay, this is what I can realistically do. This is how it impacts me, and this is how I can move forward with my life and be able to do the things I want to do.’”

To make ALS livable, we need to find ways to reduce the harms of living with ALS, which means we need to see quicker diagnoses.

“If we think about the goal of making ALS a livable disease by 2030, one of the steppingstones is definitely the top of that funnel of trying to get the patients to a diagnosis much, much, much faster,” Leavitt said.

But the active participation of the community will be essential, from “the beginning of study design around the AI algorithm to interpretation of the results to patient education material design,” according to Leavitt.

“I think people living with the disease will play a very, very important role in AI truly recognizing its potential for ALS.”

Miss this ALS Nexus session? You can still watch it and all the others on-demand. Find out how at alsnexus.org. And don’t forget to save the date! ALS Nexus 2025 is August 11–14 in Dallas, Texas.

To get the latest updates in ALS research, subscribe to our monthly e-newsletter, Research Matters. 

To continue to follow stories about people living with ALS in the community and learn more about the disease, subscribe to receive our weekly blogs in your inbox HERE or follow us at als.org/blog. 

Subscribe