BACKGROUND:
Stroke, defined by NICE as a vascular syndrome characterised by rapid-onset disturbance of cerebral function, affects approximately 100,000 people each year, with over 1.4 million survivors in the UK. Cognitive impairment is a common sequela of stroke that impedes recovery and requires systematic assessment as part of rehabilitation. The Montreal Cognitive Assessment (MoCA) is widely used for this purpose but has several limitations, including a disregard for physical disability, the need for trained staff and a lack of an integrated mood assessment. CognoStroke, a component of the CognoSpeak digital assessment platform, was developed to address these limitations. It replicates the cognitive domains assessed by the MoCA, while incorporating mood screening using the GAD-7 and PHQ-9, removing the need for staff presence and accommodating individual patient needs. CognoStroke provides automated scoring and longitudinal performance tracking, via the use of artificial intelligence, supporting clinical decision-making and subsequently, aiding stroke rehabilitation.
METHODS:
Patients with confirmed strokes, TIAs, or stroke mimics were recruited from inpatient and outpatient settings (n=28). Following capacity assessment and consent, participants completed both CognoStroke and MoCA assessments, either in person or remotely, and provided feedback on usability and attitudes towards artificial intelligence.
RESULTS:
Preliminary findings indicate high user acceptability, with 96.3% of participants reporting that they liked or felt neutrally towards speaking to the digital agent. Signs of severe anxiety were detected in 7.1% of participants and signs of moderately severe depression were detected in 10.7% of participants, suggesting that CognoStroke may be a more feasible and holistic digital alternative for post-stroke cognitive assessment. Challenges remain regarding accessibility and technological literacy amongst stroke survivors, with only 33.3% of participants completing the CognoStroke assessment independently. However, ongoing system adaptations are aiming to enhance user autonomy and inclusivity.
CONCLUSION:
CognoStroke shows strong potential as a scalable and clinically meaningful digital cognitive assessment tool for stroke survivors, pending further validation to support clinical implementation.