Can artificial intelligence improve the diagnosis and prognosis of disorders of consciousness? A scoping review.

BACKGROUND

Artificial intelligence (AI), in the form of machine learning (ML) or deep learning (DL) models, can aid clinicians in the diagnostic process and/or in the prognosis of critically medical conditions, as for patients with a disorder of consciousness (DoC), in which both aspects are particularly challenging. DoC is a category of neurological impairments that are mainly caused by severe acquired brain injury, like ischemic or hemorrhagic strokes or traumatic injuries. The aim of this scoping review is to map the literature on the role of ML and DL in the field of diagnosis and prognosis of DoCs.

MATERIALS AND METHODS

A scoping search, started from 3rd October 2024, was conducted for all peer-reviewed articles published from 2000 to 2024, using the following databases: PubMed, Embase, Scopus and Cochrane Library.

RESULTS

We found a total of 49,417 articles. After duplicate removal and title/abstract screening, 613 articles met the inclusion criteria, but 592 articles were excluded after full-text review. Therefore, only 21 studies involving DoC subjects were included in the review synthesis.

CONCLUSION

Advancing AI in the field of DoC requires standardized data protocols and consideration of demographic variations. AI could enhance diagnosis, prognosis, and differentiation between states like unresponsive wakefulness syndrome (UWS) and minimally conscious state (MCS). Additionally, AI-based applications personalize rehabilitation by identifying key recovery factors, optimizing patient outcomes.