Early diagnosis of Alzheimer's dementia with the artificial intelligence‐based Integrated Cognitive Assessment
Bibliographic record
Abstract
Abstract Background We have developed the Integrated Cognitive Assessment (ICA), a 5‐minute, self‐administered, computerised test that is independent of language, cultural background and education and aims at screening for cognitive impairment in a way that can simplify and accelerate the diagnosis of Alzheimer's Dementia (AD) and Mild Cognitive Impairment (MCI). The ICA utilises artificial intelligence to analyse high‐dimensional clinical and demographic data. Method We carried out head‐to‐head studies comparing classification performance of the ICA with widely used cognitive assessments (MoCA and ACE) in participants with MCI and mild AD. The ICA test measures patterns of reaction time and categorisation accuracy which are utilised by an AI engine, alongside demographic data, to provide a predictive score about participant’s cognitive status. We also investigated the use of a deep (50 layers) neural network to extract informative features from the ICA test response patterns. Result On a population of 200 participants (84 healthy, 68 MCI, 48 mild AD), the ICA achieved an area under the ROC accuracy of 91% in distinguishing between healthy and impaired (MCI and mild AD) participants. In comparison MoCA achieved an AUC of 82%, and ACE 84%. Utilising the deep learning network for automatic feature extraction significantly improved the specificity and sensitivity compared to only using a linear classifier. The ICA Spearman correlation of 0.67 (p‐value <0.0001) with MoCA, and 0.73 (p‐value<0.0001) with ACE establishes convergent validity with these cognitive tests. ICA results were not biased by participants level of education (i.e. no significant correlation), whereas MoCA and ACE had correlations of 0.31 (p<0.0001) and 0.31 (p<0.001) respectively with the level of education in the same set of subjects. Conclusion The ICA can support clinicians by aiding accurate diagnosis of MCI and AD and is appropriate for large‐scale screening of cognitive impairment. The ICA has advantages over MoCA and ACE because of its shorter duration, automatic scoring and potential for medical record or research database integration. ICA’s AI engine is able to learn from additional data and utilise deep learning, further improving the predictive power of the ICA test.
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 | 0.002 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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".