AI-enabled digital health tech catches mild cognitive impairment early

Dr. Alvaro Pascual-Leone, a practicing neurologist and cofounder of Linus Health, offers an in-depth look at his company's digital replacement for a longtime paper test that helps assess risk of dementia and stroke.
By Bill Siwicki
11:27 AM

Dr. Alvaro Pascual-Leone (far right) and some of the Linus Health team at the Alzheimer's Association International Conference 2023 in Amsterdam

Photo: Dr. Alvaro Pascual-Leone

There's been a flurry of attention around a new study in Frontiers in Neurology validating that the Brain Care Score, or BCS, may help in assessing one's risk of developing dementia or having a stroke as they age.

But what happens next?

Dr. Alvaro Pascual-Leone, a practicing neurologist, chief medical officer and cofounder of Linus Health, and a professor in neurology at Harvard Medical School, has a plan. It's a testing regime in digital form that doctors can hand to patients on iPads, with a digital clock test being a central component.

Linus Health's next-generation digital cognitive assessment platform supports early detection and intervention in brain health. It can be used as a screening tool anywhere within a health system – within the emergency department or at a PCP or neurologist office – to flag mild cognitive impairment early for earlier interventions and better outcomes, Pascual-Leone said.

We interviewed Pascual-Leone to get the lowdown on this digital testing regimen and what he hopes it will achieve.

Q. What was the problem you were trying to address when you came up with the iPad-based digital screening for mild cognitive impairment?

A. Brain-related disability is the No. 1 cause of disability, more so than cancer and cardiovascular disease combined. The disability caused by Alzheimer's and other forms of dementia is the No. 1 disability because it threatens to not just take away our mobility or vitality, but the core essence of who we are.

Yet despite these realities, our status quo approach to brain health, to prevent this devastating disability, is amply proven to fail because it tends to be very reactive. As patients and providers, we typically wait until we are overwhelmed by symptoms or problems, or our loved ones are overwhelmed, and it is too late to do anything meaningful.

When we do seek help, it typically is in the form of being referred to a specialist, of which there is a tremendous shortage. Even when a preventive approach is taken, either due to a family history of dementia or general concern, it often is without evidence-based guidance personalized to our health profile and thus it fails to be as effective as it could be.

What Linus Health envisioned is taking a lesson from motorsports and implementing the idea of the "pit stop." It wasn't until the 1960s that the Wood Brothers Racing team standardized the "pit stop" process to prevent breakdowns during races, even though the sport had been popular for decades.

Brain health needed a similar preventive approach involving anticipating problems and receiving periodic assessments in a very targeted, focused and actionable way.

Like the modern pit stop, though, brain health evaluation ultimately needs to be quick, at the right time, and meaningful action taken when early signs of mild cognitive impairment (MCI) are detected. While an assessment alone could be helpful, primary care physicians need much more data and tools to set patients and families on an effective and personalized care pathway.

We determined early on that AI-enabled digital health technology was the best method for generating and analyzing an abundance of such data in a short amount of time.

Q. How does the digital screening for mild cognitive impairment work? How is it different from the manual version?

A. The platform is based on the paper Clock Drawing Test, which involves the provider asking the patient to draw a clock face on a piece of paper, including all the numbers and hands arranged to a specific time. Before our technology was developed, the test was well-established by evidence as a reliable indicator of cognitive impairment.

Despite the simplicity, there are many neurocognitive activities involved, tapping into our memory, planning, executive function, and visual-spatial abilities that all need to come together to understand the instructions and perform the task.

The greater value of the test, however, comes from a trained clinician watching the patient perform the test, including the hesitations, doubts and corrections we make, all of which are even more informative than the final drawing. This is known as the Boston Process Approach – assessing the way in which we go about solving a task, rather than simply evaluating the end product.

Yet capturing and interpreting key moments from the drawing process requires expertise and takes time, which most primary care physicians do not have, while the wait time for an appointment with a neuropsychologist, a neurologist or other brain health specialist usually is many months.

Using an iPad and digital pen, the technology can capture and interpret hundreds of metrics from the patient's digital clock drawing in less than three minutes, far more meaningful information than could ever be gathered from a paper-based test.

The technology can detect subtle signals such as variations in hand movement, pressure on the screen, placement of numbers, how the circle is drawn and other relevant findings. After the test, providers and patients receive a green, yellow or red score.

If cognitive impairment is demonstrated, the technology delivers an evidence-based action plan, including recommendations about next steps, which typically include personalized lifestyle interventions involving exercise, diet, cognitive skill-building activities, a hearing test, recommendations for further laboratory tests or brain imaging if indicated, etc. A referral to a specialist may be recommended depending on the findings but is not always necessary for all patients.

Q. What results have you seen to date where hospitals, health systems and group practices are using the iPad-based technology? What are the outcomes in process and patients?

A. The digital clock test builds on many decades of extensive experience with the paper-and-pencil version of the same task. Digital clock test technology was developed and extensively studied at Lahey Hospital & Medical Center, which is affiliated with Tufts Medical School, and MIT.

Findings with the digital clock test have appeared in more than 20 peer-reviewed, scientific publications and new studies continue to be published every year. At the beginning of 2024, for example, in a paper published in Alzheimer's Research and Therapy, researchers concluded the digital clock test outperformed the Mini-Mental State Examination (a paper test) in detecting early signs of MCI.

In these studies, researchers and providers have found that although the tests are very short and easy to implement in a fast-paced outpatient setting, they are exceedingly sensitive due to the abundance of data captured and analyzed.

Research has shown, in fact, that the test is more sensitive than a standard battery of neurocognitive tests that typically requires much more time to perform, over many hours, in some cases.

Moreover, providers have found the technology can detect, with great precision and sensitivity, early signs of MCI across adults older than 65, but also younger age groups with diverse racial and cultural backgrounds.

Due to the high sensitivity and short time duration, it has been a highly accepted diagnostic tool by clinicians, who also appreciate that testing is easily repeatable to compare results over time. Because no significant training is required, medical assistants or technicians can supervise the test while the physician can then easily review results with the patient at the point of care.

Results are particularly compelling to providers as they are a valuable aid to them in supporting an accurate diagnostic decision, as well as providing a highly personalized prognosis and care plan.

Thanks to the fact that the Linus Health platform captures multiple types of signal and metrics, for example, the patient's pressure on the pen, how they are holding the pen, their drawing consistency and pauses, the amount of pressure put with the pen on the tablet, the size, position and orientation of the various drawn elements, etc., it is then possible to apply machine learning and generate different algorithms to assess different aspects of brain health all in parallel, known as "a multiplex of insights."

For example, in addition to determining if a patient is cognitively impaired, the analytics technology can quantify the patient's risk of progressing to deeper dementia and distinguish between different conditions, which provides primary care physicians without specialized training or experience an incredibly powerful triage tool.

That insight can support a more fully informed referral to a neurologist or other specialist so more targeted and relevant testing can be performed. The care pathway is then more efficient, helping rule out if the patient's cognitive impairment is caused by a thyroid condition, for example, or sleep apnea or major depressive disorder.

This saves time for providers, but also for patients who are so emotionally invested and motivated to find an accurate diagnosis and effective care plan to slow the progression of their condition.

Q. What do you think the future looks like for digital health in the treatment of cognitive impairment?

A. A prevention-oriented "pit stop model" that I referenced earlier is where brain health management is headed. Preventive care has been the mantra for cardiovascular conditions for decades. As a result, while heart disease is still the No. 1 cause of death, we are much better as providers at managing conditions such as high blood pressure, hyperlipidemia, heart failure and others.

The same, however, cannot be said about neurocognitive conditions, which are still approached far too reactively and tend to make pharmacologic, behavioral and lifestyle interventions more difficult to implement and less effective.

Digital health technology will shift the management approach to dementia and brain health to a preventive care paradigm that enables primary care physicians to lead that charge. With precise and highly personalized digital health tools, we can help patients and caregivers manage their condition, but also define their brain health goals earlier to tailor interventions.

This is an important factor because patients' reasons for slowing their condition's progression are going to vary, as will the activities of daily life they want to continue to participate in and enjoy for as long as possible.

By approaching the condition from a personalized and preventive posture, we can deliver the interventions, guidance, coaching and treatments that will enable patients to stay in their race for as long as possible and continue to enrich their lives and the lives of loved ones and pursue the activities they find most fulfilling.

Follow Bill's HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.

Want to get more stories like this one? Get daily news updates from Healthcare IT News.
Your subscription has been saved.
Something went wrong. Please try again.