Will AI transform how we fight dementia?

How neural networks are helping neurologists with the diagnosis and treatment of neurological disorders

According to British Alzheimer’s Disease International, there are almost ten million new cases of dementia each year worldwide, implying one new case every few seconds. I can see it in my own environment; many of my friends’ parents and my parents friends have sharply declined in cognitive ability, with Alzheimer and Parkinson’s being the major cause. They tell me heartwrenching stories how their parents fail to recognize them or how they need support in the most basic areas of personal care. Janet Barberg shares her daily experiences online. In her words “My husband and I have been caring for my 92 year old mother for over five years. It has taken me almost that long to even admit and verbalize that my mother has dementia. I always would just say that she was forgetful and then as time went on I added the word confused.”

For a long time, the brain was considered a mysterious black box.

We have made some extraordinary progress in medicine the past decades. Cardiovascalur disease, the main cause of mortality, has seen dramatically improved treatment outcomes in the past 30 years. With AI-enabled diagnostic tools and new immunotherapies on the horizon we expect better chances for those suffering from cancer in the near future. For a long time, the brain was considered a mysterious black box. It is the only organ that we cannot live without and that can not be replaced. While neuroscientists are gradually uncovering more and more about the workings of the brain, neurology remains one of the most challenging disciplines in medicine. Technologies such as MR imaging to identify subtle changes in the brain tissue and structure, fMRI imaging to measure changes in brain activity, and EEG to measure electrical activity have given clinicians many new insights into what is happening in brain health and disease.

At Philips we are working with researchers to stimulate the brain using electrodes that are placed on the skull, applying Artificial Intelligence (AI) to personalize treatment for patients with severe epilepsy. These implantable deep brain stimulation devices can help control tremors or pain.

To enable targeted therapy for patients, more precise diagnosis is required. In some types of dementia, where there is a long period between contracting the disease and the first symptoms, an early diagnosis could be critical to initiate treatment as early as possible. In patients with various neurological and mental conditions we are exploring the fusion of physiological measurements, contextual observations, and measurements with wearables, apps and AI to identify patterns and correlations leading to an early, confident diagnosis. The next step could be early intervention or even prevention in risk groups, when connections are discovered between lifestyle and cognitive impairment. These solutions could positively impact millions of people.

Although preventative care in neurology is still emerging, we already see areas where digital technologies can help to make great strides. An example is how the latest generation of digital MR imaging technologies is helping research in the quantification of iron accumulation associated with many neuro-degenerative diseases. We are also using AI to spot early indication of plaque in the brain and measure brain atrophy (yes your brain is shrinking!) compared to normal.

 From subjective to quantified results

There are other examples of how digital technologies could positively impact the quality of care for people living with neurological disorders. Sometimes there are subtle characteristics of thinking and behavior that reveal the onset of such diseases. Today, neurologists commonly either refer a person for a full neuro-psychological assessment, which is both time consuming and costly, or ask the person to perform a cognitive task – such as copying a drawing, remembering a series of numbers, or solving a problem – and observe how they do it. Evaluation and understanding of cognitive function is critical for diagnosis, but there are many barriers when trying to assess patients. Another challenge involves deciding on the best-fit therapy for each patient. Depending on the disorder, a clinician will combine cognitive function with other data sources, such as neuro-imaging and other tests, and choose the appropriate treatment path. This can often mean recommending medications or some form of cognitive rehabilitation or behavioral therapy. In many cases, however, treatment guidelines lack clarity. It can also be extremely difficult to spot patterns of cognitive deficits, especially over longer periods of time, and to use these to assess the effectiveness of treatment.

Fifty percent of neurologists think the average wait time for a cognitive assessment has a negative impact on diagnosis and treatment.

With the increasing demand for mental care, neurologists lack time and resources to adequately test for impairment and make objective and informed recommendations. Fifty percent of neurologists think the average wait time for a cognitive assessment has a negative impact on diagnosis and treatment*.

New AI-powered digital cognitive assessment platform

Working with neuropsychologists and neurologists, Philips is currently developing a digital cognitive assessment solution, IntelliSpace Cognition, which combines a range of well-established tests, provides automated scoring, translates test scores to cognitive domains, and offers basic interpretation. The platform uses Artificial Intelligence to provide quantified, longitudinal insights from data-rich assessments. A primary goal of this solution is to facilitiate robust and reliable cognitive assessment with far greater efficiency and scale than traditional methods. The emergence of digital platforms that combine the best of traditional neuropsychological assessment with cutting-edge automation and AI has the potential to meaningfully transform neurologists’ diagnoses and treatment plans for improved patient care.  

The brain is the next frontier for healthcare. With the fusion of medical modalities, insights gleaned from wearables, EMRs, apps and the application of AI, we will make great strides in the coming years to better understand and diagnose neurological disorders and find new therapies that will lead to better mental health and hopefully allow many of us to live with dignity in the golden years of our lives.

*From an online survey of 75 US-based neurologists; commissioned by Philips and conducted by MarkTech Group.

*Product is under development and not yet available in any market.