Artificial Intelligence with Brains. The possibilities…

This week I attended the International Progressive MS Alliance webcast, with a focus on how blood tests and artificial intelligence (AI) in MRIs could predict MS progression. Hosted by Trishna Bharadia, International Patient Advocate, who I had the pleasure of meeting at MS Session in Prague; Dr.Robert Fox, neurologist and Vice Chair for Research of the Neurological Institute at the Cleveland Clinic, US; Professor Douglas Arnold, Neurology at McGill University, Canada and; Professor Tal Arbel, Electrical and Computer Engineering at McGill University, Canada.

With my MS taking quite a progressive hit this past year or so, I am extremely keen (more so than before) in keeping myself and others updated on the many research studies, and the conferences/webinars etc. that are happening surrounding progressive MS in particular. It has become even more apparent that there isn’t a quick way in detecting the effectiveness on whether a therapy is working or not, since MS progression can occur slowly. I for one had my second Alemtuzumab (Lemtrada) infusion course in April 2019, but the progression with my MS has been rather evident since then with a physical and cognitive decline, despite there being no new lesions on my recent MRI scan. 

This is all down to biomarkers, which tell you things earlier than what you may see clinically. For example, they can measures/highlight what’s going wrong in the body with various things; i.e. blood count, oxygen levels, blood pressure etc. They are warning signs. In RRMS the new lesions on an MRI is used as a biomarker, an indication that something new is occurring. This logic can be applied as treatment markers if you will, to indicate the response to a treatment. If there aren’t any new lesions highlighting on an MRI then this is used as an indicator that the treatment is working in regards to no new activity/relapses and that the treatment is effectively doing its job in that regard, but what about progressive MS? An MRI shows the treatment efficiency of the drugs with bio markers (lesions) in RRMS, but there are no comparable biomarkers for progressive MS, and this is what is being researched at the moment with AI being developed with biomarkers on MRIs in PMS. 

There have been a multitude of studies and publications focusing on AI being used alongside MRIs to predict MS progression, which are on-going. Take a look at the brain images below and the corresponding ages of each one. You will notice the healthy brain in image 1, but compare that to image 3 which is the brain of a similar age who has had RRMS for 7 years. Notice that the brain predicted age is that of a healthy 40 year old? +9 years? Alongside that, take a look at image 4 of an individual with SPMS who would have previously had RRMS before transitioning to SPMS – what if we are able to analyse the brain in Image 4 on where the lesions are, the timescales at which EDSS score was reached etc, and apply this to the brain of image 3 to predict where the progression is likely to head and intervene at that point? Then what if we are able to develop a treatment for the predicted progression of SPMS and PPMS? Just.imagine. 

(Cole, Ann Neurol 2020)

Brains shrink for everyone as we get older, healthy or otherwise, but brain shrinkage also occurs when nerve cells die – as is the case with MS. What if we can predict the course an ‘MS brain’ is likely to take in regards to progression and the shrinkage that is occurring in these images, for example. What if we can interfere at the point of when a DMT is chosen on whether the treatment can cater for the predicted disease progression of each individual? We can apply this to my case where I do not have any new activity, but I am progressing with my previous symptoms which do not highlight on MRIs. What if we analyse where exactly the previous lesions are on my brain, and correlate this to the predicted progression path through AI? With this we will not only know whether a treatment is predicted to be effective or not before it is even taken, but where an individual’s progression is headed. 

Of course, there are still questions that need to be answered before we reach the same level with RRMS treatment compared to progressive, which is very limited at this point. What is truly driving MS progression? What’s not working? What exactly is declining? What features of the MRI are associated to the progression? We need to understand that better, then the biomarkers will be useful to get to the core of understanding the present lack of clarity in PMS (Fox, Rebert 2020).

Off the back of this, I have requested my most recent MRI scan (Oct ’20) with lesion details to have good nose at.


One thought on “Artificial Intelligence with Brains. The possibilities…

Add yours

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at

Up ↑

%d bloggers like this: