Feasibility of longitudinal automated cognitive assessment in the stroke pathway
Bibliographic record
Abstract
Abstract Background Stroke, post‐stroke dementia and post‐stroke cognitive impairment prevalence is rising significantly, placing an increasing burden on healthcare systems. Standardised pen and paper tools for cognitive assessment require clinical time. There is increasing research into the use of automated cognitive assessment. CognoSpeak (https://cognospeak.github.io/website/) is a fully automated cognitive assessment tool based in language an interaction. A virtual clinician asking questions and listening to responses using automatic speech recognition and Machine Learning algorithms. Methods Patients who passed the eligibility criteria of having recent acute stroke/TIA, and do not have pre‐existing dementia, severe aphasia, deafness, or too medically unwell were approached to consent to speak to CognoSpeak. Participants completed the CognoSpeak assessment on the ward or at home via a web‐version, as well as MOCA and GAD and PHQ9. Results Recruitment started December 2020 and we approached 950 people and recruited 114 participants, with 20 having 6 months follow‐up and 5 12 months follow‐up. The mean age was 61.77 (SD 15.07). Most participants were recruited from Sheffield 71 (68%).The majority had ischemic stroke (67%) & mild stroke according to the NHISS (63%). The most common risk factor was hypertension (19%), and the majority of participants had no symptoms of disability before their stroke (22%). GAD and PHQ9 scores were collected. Conclusion It is feasible to complete automated cognitive assessment with the stroke population. There remains barriers in utilising technology within this acute and especially longitudinal cognitive assessment in stroke survivors, where rates of anxiety in hospital and at follow‐up appointments may reflect rapid discharge and changes in the stroke pathway during the initial covid pandemic.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.042 | 0.074 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.004 | 0.004 |
| Open science | 0.002 | 0.004 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.006 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".