Treating TBI patients in the intensive care unit (ICU) is challenging for main three reasons, says Rahul Raj.
“First of all, the patient is usually unconscious. Second, there is no objective way to measure the severity of the brain injury. Third, we don’t objectively know how different treatments impact the prognosis of the patients.”
The recognised research paper shows a promising start for future use of machine-learning based solutions in intensive care. It is the world’s first algorithm based on machine-learning, that can help clinicians to quantify the dynamic prognosis of patients with TBI. In practice, this means that there will be a scale of 0 to 100 to describe the probability of death caused by the brain injury.
“With the algorithm we can monitor the patient’s prognosis with intensive care every eight hours even though the patient is unconscious. This way we may also get new information regarding how different treatments affect the patient and assess if they are useful or perhaps harmful,” Raj says.
The work is ongoing and the model will be further tested and developed before someday hopefully it can be tested in the clinics. Rahul Raj emphasises that the code for the algorithm is available online for other researches to develop.
HUS press release in Finnish
tel. +358 50 427 0698