Could artificial intelligence predict the outcomes of patients with TBI in real time?

A team of researchers have presented an artificial intelligence-based algorithm with the potential to predict mortality in real time during intensive care after severe TBI.

Like Comment

In a collaboration project between Helsinki University Hospital (HUS), Kuopio University Hospital and Turku University Hospital (all Finland), a team of researchers have presented the first artificial intelligence (AI) based algorithm that has the potential to assist in treating patients with severe TBI in intensive care units (ICUs).

Patients with the most severe cases of TBI are usually treated in ICUs, however, despite the high-quality care, recent observational studies have reported mortality rates of approximately 30%.

Patients who suffer from severe TBI are unconscious, therefore, it is a challenge to accurately monitor their condition. In ICUs many tens of variables, such as intercranial pressure and mean arterial pressure, are continuously monitored to assess the patient’s condition.

One variable alone could yield hundreds of thousands of data points per day, making it impossible for ICU staff to fully analyze. This is what motivated the team of researchers to develop an AI-based algorithm that could evaluate all the collected data, assisting clinicians in treating TBI patient in an ICU.

The algorithm has the potential to predict the outcomes of individual patients, provide an objective assessment of the patient’s current condition and how it changes during treatment.

“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,” commented study author and Adjunct Professor of Experimental Neurosurgery, Rahul Raj (HUS).

In the study, the algorithm was reported to be able to predict the probability of patient mortality within 30 days with an accuracy of 80-85%.

“We have developed two separate algorithms. The first algorithm 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…Still, the accuracy of both algorithms is surprisingly good,” explained Eetu Pursiainen (HUS), study author and Data Scientist.

The positive results from this study suggest that these algorithms could aid in clinical decision making in the future and encourage clinicians to make more data-driven treatment decisions, which could potentially improve quality of care.

“We think that it is important [to] act ethically and share our algorithms openly and free of charge for further development, both nationally and internationally,” concluded Miikka Korja, Chair of the HUS Artificial Intelligence Steering Group and Adjunct Professor of Neurosurgery at the University of Helsinki.

Source: Raj R, Luostarinen T, Pursianinen E et al. Machine learning-based dynamic mortality prediction after traumatic brain injury. Sci. Rep. doi:10.1038/s41598-019-53889-6 (2019) (Epub ahead of print);

Go to the profile of Kimberley Ndungu

Kimberley Ndungu

Editor, Concussion Zone

I am the Editor of Concussion Zone, so feel free to get in contact if you have any queries or comments.

No comments yet.