Most severe TBIs are treated in intensive care units (ICUs), but in spite of proper and high- quality care, about one in three patients dies, researchers behind the study say.
Their paper aims to use AI to tackle the challenge of accurately monitoring TBI patients in intensive care when they are unconscious.
The Finnish researchers say: “In the ICU, many tens of variables are continuously monitored – such as intracranial pressure, mean arterial pressure and cerebral perfusion pressure – that indirectly give information regarding the condition of the patient.
“However, only one variable, such as intracranial pressure, may yield hundreds of thousands of data points per day. Thus, it is impossible for the human brain to comprehend the resulting millions of daily collected data points from all monitored data.”
They set out to develop an AI-based algorithm that could help to predict the outcome of the individual patient and give objective data regarding the condition and prognosis of the patient and how it changes during treatment.
Rahul Raj, an author of the paper and a professor of experimental neurosurgery, says: “A dynamic prognostic model like this has not been presented before.
Although this is a proof-of-concept and it will still take some time before we can implement algorithms like this into daily clinical practice, our study reflects how and into what direction modern intensive care is evolving.”
The algorithms they developed can predict the probability of the patient dying within 30- days with accuracy of 80 to 85 per cent.
They say: “We have developed two separate algorithms. The first is simpler and is based only upon objective monitor data.
“The second algorithm is slightly more complex and includes data regarding the level of consciousness, measured by the widely used Glasgow Coma Scale score.
“As expected, the accuracy of the more complex algorithm is slightly better than for the simpler algorithm. Still, the accuracy of both algorithms is surprisingly good, considering that the simpler model is based upon only three main variables and the more complex upon five main variables.”
The algorithms are yet to be validated in national and international external datasets, however. The project was a collaboration between three Finnish university hospitals: Helsinki University Hospital, Kuopio University Hospital and Turku University Hospital.