Connect with us
  • Elysium

Neuro rehab technology

“We could effectively build 66 hospitals” – Sanome CEO on game-changing digital twin tech



SR Times talks to Benedikt Von Thüngen, CEO of health tech company Sanome, on how its innovations could power the future of stroke rehabilitation.

Sanome is developing a ‘human digital twin’ which uses AI in order to create an early warning system using biomarkers. For stroke, this could predict rehab outcomes in patients.

SR Times: What was the inspiration behind Sanome?

Benedikt Von Thüngen: “My personal background is biomedical sciences, with a further interest in healthcare. I’ve always been fascinated in what can we actually do in healthcare.

“My second company was one of the first companies looking at digital health. Back in 2013 I was getting really excited by apps, but the fundamental challenge of healthcare is that there is a ginormous amount of data that gets generated in a clinical setting, however, at the time that data had barely been digitised.

“So, I left the healthcare market to focus on my speech recognition company, as I waited for the time when the healthcare data became increasingly digitised.

“In terms of the inspiration for Sanome, if you look at healthcare and the supply and demand needed, we are seeing a major increase in the UK and on a global basis. 

“We can argue about hiring, building more hospitals, hiring more doctors, nurses and carers, but the money is just not there anymore.

“However, at the same time, the demand for healthcare is growing, people are living longer and people are getting sicker. 

“How can we leverage the data that is increasingly available there to bring those two curves closer together?

“The only way to do that is by using data to predict patient outcomes, or predict patient deterioration effectively, [as an] early warning system. It’s all in the data. When we get sick, you normally know a day or two in advance. You feel a little bit groggy, or have a bit of a headache, so the data signals are there, but, we don’t do anything about it.

“When you apply that to more complex disease, the signals are there but we only act really, really late. So going back to that big curve, if we could make all of that [happen] earlier, it becomes cheaper to treat the individual, bringing supply and demand closer together.

“So that has been the inspiration for Sanome, saying ‘right lets take all the data that’s now increasingly available [through] AI, ML (machine learning) algorithms that have been developed over the last decade to really utilise that to build early warning system that detect health deterioration earlier’.”

What is a human digital twin?

“A human digital twin, ultimately, is an early warning system.

“So you’re predicting simulating where patients are going to end up over time to then give you something to action on today. And that’s what an early warning system is all about.”

How do these biomarkers work and allow predictions to be made?

“If we relate specifically to stroke, we’re interested in two things. The rehab element, where is the patient going to end up in three months, six months or 12 months?

“Are they going to regain consciousness, are they going to regain continence? Are they going to regain mobility? Are they going to be able to feed themselves? It’s very much kind of black box. It’s guesswork. So, that is one scenario, it’s the outcome.

“The second thing we care about is what about deteriorations [such as] a second stroke, a heart attack or an infection. And this is really hard because these patient groups often can’t communicate.

“We’re worried deterioration and outcomes. What we do know is that you can typically see trends in the data.

“AI algorithms are really good at detecting the subtlest changes, that could give us an indication of where it goes.

“So if we take a human digital twin, we take a stroke of rehab patient and that data feeds into this model, the model then says based on having seen X amount of exams before, I think there’s a fairly good likelihood that this patient is going to end up here.

“Care intensity is something that we really care about, particularly in rehab patients, because what we want to see why does music therapy help?

“Why does physiotherapy help? Should we do it once a week? Should we do it five times a week? So that kind of care intensity, you can start seeing if the prediction accuracy improves, as you put the certain therapy events in, or take them away again?

“So that’s what the digital twin does is kind of that feedback loop, there’s new bit of data coming in, does it improve? Or does it change the prediction?”

Could Sanome’s technology help to save public health systems including the NHS significant funds?

“Avoidable events in the UK cost the NHS £8bn pounds. It is estimated that somewhere between 10 and 15 per cent of all hospital beds in the UK are occupied because of one of these entirely avoidable health conditions and the chance of a patient getting one of these avoidable events, particularly if they’re like a rehab centre is 80 to 90 per cent.

“It’s just going to happen. You’re going to get a UTI, you’re going to get pneumonia, or something like that is going to happen, and that has a huge health care cost, but also it clogs up the system.

“Let’s say you’re on a rehab pathway, and you have to go to the ICU, because you have pneumonia, that kind of destroys all of that rehab. So, you basically go back again and restart everything. That indirect cost is estimated to be in the billions as well.

“The first step we’re doing is predicting or detecting these events sooner.

“[Former UK Prime Minister] Boris Johnson claimed that he was going to build 40 hospitals. If, and this is a big if, we can get this early warning system that we’re building into every hospital in the UK, it reduces the number of beds, because you’re reducing the number of incidents, we effectively build 66 hospitals.

“What it really means from a patient perspective is you cannot stop someone getting an infection, but you can, if you detect it on day one or day two. That’s the difference between going to the ICU or not.

“That’s the difference between dying or not. It means staying in the room that you’re familiar with, and continuing your treatment, but just receiving an oral antibiotic, so that cost reduction is immense.”

Find out more about Sanome here.