Study: ML algorithm helps detect traumatic intracranial hemorrhage using prehospital data


A machine learning algorithm can accurately detect traumatic intracranial hemorrhage using information collected before patients reach the hospital, according to a study published in JAMA Network Open

Researchers built a prehospital triage system using data paramedics could provide, including the patient’s age, sex, systolic blood pressure, heart rate, body temperature, respiratory rate, consciousness, pupil abnormalities, post-traumatic seizures, vomiting, hemiplegia, clinical deterioration, whether head trauma occurred from great force or pressure, and whether the patient suffered multiple injuries. 

The study analyzed electronic health records from 2,123 patients with head trauma who were transported to Tokyo Medical and Dental University Hospital from April 1, 2018, to March 31, 2021. The machine learning model detected traumatic intracranial hemorrhage with a sensitivity of 74% and a specificity of 75% using prehospital information.

Comparatively, a prediction model using the National Institute for Health and Care Excellence (NICE) guidelines, calculated after consulting with physicians, had a sensitivity of 72% and a specificity of 73%, which was not statistically different from the prehospital model. 

“Although conventional screening tools require examination by a physician, our proposed models require only pre-transportation patient information, which can be easily obtained,” the study’s authors wrote. 

“The results suggest that our proposed prediction models may be useful for constructing a triage system that can be used to assess the optimal institution to which a patient with a head injury should be transported. Further validation with prospective and multicenter data sets is needed.”


Researchers said assessing head trauma in the field could improve outcomes for patients. The current system for head trauma requires paramedics to bring patients to the hospital if they decide it’s necessary, where a doctor would evaluate whether a patient needs a CT scan. After a scan, the patient may need to be transported to another hospital.

Adding field triage could allow ambulances to bring patients to the best site for care first, decreasing time to treatment. 

“As the functional outcomes of patients with head injury worsen when their transportation is delayed, the transport time in step three should be reduced by constructing a reliable field triage tool,” the researchers wrote. 


As artificial intelligence use expands in healthcare, experts and studies have noted the importance of monitoring for bias, which could worsen existing health inequities. 

AI developers also need to conduct thorough testing to ensure the model works in all environments. Researchers in this head trauma study noted this is one limitation of their study, because it focused on a single site in Japan.

“Because this was a single-center study and included only patients who were hospitalized and underwent head CT, our data set may not represent the general population of patients with head trauma,” they wrote.

“In addition, we suggest that our model may be underestimating patients at high risk, based on the calibration plot. To apply our model to clinical practice, we should verify the predictive accuracy using a prospective external validation set and investigate the optimal cutoff value.”


Source link

Leave a Reply

Your email address will not be published. Required fields are marked *