Educational content on VJNeurology is intended for healthcare professionals only. By visiting this website and accessing this information you confirm that you are a healthcare professional.

Share this video  

ESOC 2022 | Using machine learning to detect stroke cases during emergency services calls

Helle Collatz Christensen, MD, PhD, University of Copenhagen, Copenhagen, discusses the implementation of a machine learning model trained to identify stroke and alert dispatchers during emergency calls. Recorded stroke calls made to the emergency services in Copenhagen were used in conjunction with non-stroke calls to train the machine learning model to distinguish potential stroke cases. The algorithm, which uses complex speech recognition techniques, can then be applied to live calls to alert the dispatcher when stroke is suspected. A similar system for out-of-hospital cardiac arrest detection has been successfully implemented across Denmark. Ongoing work aims to tackle current limitations of the software which discriminates against certain pronunciations and intonations, such as foreign accents or panicked tones of voice. The scheme is starting to be implemented across Denmark as a decision support tool for emergency medical dispatchers. This interview took place at European Stroke Organisation Conference 2022 in Lyon, France.