Health and Wellness - AI iPad tech to transform dementia detection

Profile picture for user catheverett By Cath Everett June 4, 2018
Summary:
Dementia is a global concern that's only going to get worse. Early detection is crucial and tech can help.

Cognetivity dementia ipad
Dementia is one of the biggest healthcare challenges of the 21st century. Some 50 million people around the world suffer from the condition, which is a collective name for more than 100 syndromes that affect the brain and impact memory thinking, behaviour and emotion.

Moreover, nearly 10 million new cases appear each year and, as a result, dementia is expected to cost the global economy a vast $1 trillion this year, doubling to $2 trillion by 2030.

But one of the key challenges posed by the condition relates to diagnosis. According to Alzheimer’s Disease International’s ‘From plan to impact’ report, only about half of all people living with dementia in advanced economies such as the UK and US receive a diagnosis, while in the developing world the figure plummets to a mere one in 10.

But even those that are diagnosed generally receive the verdict late in the day, which means that by the time they do, they are already disabled and dependent on family and friends. This situation is not only difficult for carers but also very expensive for healthcare systems around the world - not least because all too frequently the wrong people are referred for help.

While the ‘worried well’ tend to request testing if they lose their keys once too often, true sufferers are prone to try to ignore, and even hide, their symptoms, which simply means they do not access the support they need.

A key issue here is the current lack of dementia screening tools available. Diagnosis itself, on the other hand, is based on costly, time-consuming, pen-and-paper memory tests that are not always particularly sensitive.

Nonetheless, early diagnosis is vital. Sina Habibi, co-founder and Chief Executive of Cognetivity, explains:

The topology of the disease starts 25 to 30 years before any symptoms of memory loss appear, so our aim is to make a diagnosis 10 to 15 years before then. If you can spot dementia early, you have more chance of tackling it. Today there are no drugs or treatments that can stop or reverse it, but there are treatments to help patients preserve their cognitive functions for longer. This means they are less dependent and more capable of living an ordinary life.

Founded in 2013, Cognetivity is a neurosciences firm that has set out to redefine dementia detection. it has developed a diagnostic tool, which runs on an Apple iPad and consists of a five-minute test based on “image cognition”. Patients can undertake it themselves without the need for a clinician to intervene.

The  five-minute Integrated Cognitive Assessment (ICA) tool looks for the earliest signs of impairment by testing the performance of large areas of the brain. It can also be used for remote monitoring of the progression of diseases and measuring the effectiveness of treatments.

Image cognition

They are briefly shown about 100 images that increase in complexity over time and are simply asked to respond to what they see. Meanwhile, a back-end artificial intelligence (AI) engine analyses their reactions and builds up a profile of their brain activity.

The cloud-based AI engine, which runs on Amazon Web Services, compares each individual’s responses with a database of behaviours garnered from both healthy people and those with dementia, before flagging up whether they are fine, at risk or should be referred to a specialist.

The product is currently going through its final clinical trial with 450 volunteers at the South London and Maudsley National Health Service (NHS) Foundation Trust, which is scheduled to be complete in the second quarter of 2019. The aim is to launch it initially in the UK, where it will be positioned as a “primary care triage tool”, although the US is expected to be next on the list.

The goal is to implement the AI engine into GP surgeries to enable it to trawl through patient records and identify individuals who are deemed ‘at risk’ based on factors such as their age group, length of time in education and whether they have suffered a traumatic brain injury – although Habibi is quick to point out that the system does not look at personal identifiable information.

The idea is that, if the system identifies people who are potentially at risk, GPs would request that they come to the surgery to take the Cognetivity test on an iPad provided under a new NHS initiative intended to furnish each clinic with at least one such device by 2020.

The aim is to charge each test out at a rate of £49 - although the figure will be negotiable based on the number of tests undertaken during the course of a year. The eventual goal is that such tests will form part of each patient’s annual check up.

If a patient is discovered to be at risk, GPs could then refer them on to undertake a full gamut of tests. To help diagnosis, these specialists would receive an in-depth patient report from Cognetivity rather than the summary provided to GPs. But as Habibi adds:

Other potential markets include drug discovery and development as companies here are looking for sensitive tools for recruiting patients into trials as well as undertaking micro-measurement of the impact of drugs. But the problem in the past has always been that most tests can be learned as they’re based on memory. So the idea is that by providing a more sensitive tool, research into drug discovery should stand a better chance. This means it’s a big step towards tackling the problem.