AI Could Play a Part in the Diagnosis and Treatment of Dementia

According to the WHO, around 50 million people are affected by dementia around the world.

It can cause deterioration in memory, thinking, behaviour and the ability to carry out simple daily tasks. While it mainly affects older people it isn’t a normal part of ageing.

The current thinking is the rise in cases - around 10 million a year - is because people are living longer. As a result it’s the biggest cause of disability and dependency in older people worldwide.

It’s estimated that the cost of dementia now exceeds $1 trillion globally, which makes early detection and treatment essential.

At present there is no single test to diagnose dementia. Rather it is based on a series of tests and assessments either by a GP or a memory clinic.

Patients are asked when they notice symptoms started, and whether their daily lives are being hindered in any way.

Other conditions like heart disease, diabetes and stroke all have to be ruled out as causes or are being properly managed. Similarly, medications also need to be reviewed to discount any possible side effects.

After this cognitive assessments for memory and thinking are carried out. This is followed by blood tests to rule out other conditions and a brain scan.

None of these tests can dementia alone, rather they build up a picture that allows practitioners to deliver a diagnosis with some degree of confidence. 

It’s time consuming, distressing for the patient and expensive. It also delays getting the patient the treatment they need.

That’s why a breakthrough in artificial intelligence that gives patients a diagnosis with a single brain scan is a game-changer.

AI and Dementia

Being able to rapidly diagnose dementia means getting the individual the support and treatment they need faster.

It also improves patient outcomes by cutting down on the tests and the time they have to wait to get the results. Assuming the results are conclusive enough.

This new breakthrough can not only scan and diagnose a patient the same day. There are indications it will be able to determine if the condition will remain stable, deteriorate or need urgent treatment.

Early diagnosis can mean the patient gets treated sooner, slowing down the progression of the condition. It can also help to avoid further damage, allowing the individual to live a more independent life for longer.

It’s even possible that symptoms may never occur. At least that’s the hope of the Cambridge University team behind a artificial intelligence research dedicated to tackling dementia.

Moreover, an accurate and early diagnosis allows clinicians, social services and support networks to form a scalable support plan. This will help the patient to adjust to their new reality, support positive health outcomes.

This artificial intelligence approach, developed by Professor Kourtzi is capable of identifying patterns in the scans that would be impossible to detect with the human eye. Even those of an expert neurologist. It can then compare those patterns against a vast database of scans from other dementia patients.

From there a prediction can be made about the trajectory the dementia will take and what the best course of treatment will be.

Pre-Clinical Tests and the Future

Although in its early stages, the artificial intelligence developed by Kourtzi and his team has been able to diagnose dementia years before symptoms develop. 

More impressively, the AI was accurately diagnosing dementia even when there is no obvious sign of damage in the scan.

An ongoing trial, at Addenbrooke's Hospital, Cambridge, and other memory clinics across the UK will assess whether the AI can work in a clinical setting. The intention being to score it against the conventional approaches of diagnosis.

Approximately 500 patients will participate across a range of ages, some of whom may not be showing any signs of dementia.

The results will be shared with their doctors alongside the conventional test results for comparison. Obviously where dementia has been identified they can then advise the patients what their next steps are.

One of the big advantages of this new innovation is the confidence it gives clinicians to see a way forward. It’s frustrating for all parties when they have to convey an element of doubt because certain conventional tests weren’t definitive enough.

A cut and dry diagnosis will help doctors to give patients the right information and quickly work with other clinicians and support services to put together a treatment package.

The hope is that the artificial intelligence will be used by hospitals around the world

According to the BBC, one of the first candidates to take part in the trial was a 75 year old former executive. 

Since retiring five years prior, his wife had noticed he was struggling with his memory and struggles to make himself understood. Faced with the prospect of losing their home to potentially fund care, the pair opted for the medical trial. The artificial intelligence would not only be able to give a diagnosis but allow them to plan financially. 

Another patient, aged 57, who spoke to the BBC commented he wished the technology had been available when he was diagnosed.

Not only did it take some time to gain a diagnosis but it impacted on the individual’s mental health. Primarily because he was left wondering if it was dementia or something like a tumour.

Other AI Applications

Applying artificial intelligence as a diagnosis tool is in its early stages. Using the same principles artificial intelligence can be used to identify brain cancers or tumours early on as well.

Similarly AI could be used to spot early signs of a stroke or imparied blood vessels.

It can also be used to identify mental illness and suggest treatments. Some of the more severe mental health disorders can develop in childhood. Scans at key developmental milestones may allow doctors to identify and treat these disorders before they do lasting damage.

The ability to rapidly and accurately diagnose all these conditions will make a lasting impact on the world. Patients will get life saving treatment before it’s too late, but it will also result in more cost effective treatments. Mainly because all the guesswork will be taken out of the equation. 

Earlier treatment will mean a higher recovery rate and less strain on health services. There will also be a reduced cost of sickness on the world economy.

The most exciting aspect is this is just the beginning of a new wave of AI innovation.

Round the Clock Innovation

As the applications for artificial intelligence in healthcare open up it will spark a new surge in development.

Artificial intelligence will quickly go through its next stage of growth much faster than the last and with it its uses across medicine.

At Vertrical we work with developers and engineers to create the next generation in health technology. Specifically we partner with fast growing health tech companies to help them scale up and take their product to the next level. Learn more about how we can help or get in touch with me to learn more.