Classification of patients with dementia in a late life depression cohort using structural MRI
Bibliographic record
Abstract
Abstract Background Late life depression (LLD) is a risk factor for cognitive impairment, with up to 40% of depressed older adults showing persistent cognitive deficits even after symptom remission. Over 20% of Alzheimer’s disease patients exhibit depression‐like symptoms which can represent an early marker of disease (Tsuno et Homma., 2009). We hypothesized that automated classification algorithms trained on structural MRI could predict adjudicated diagnosis of cognitive impairment (CI) in this cohort. Method Our study evaluated 295 patients with LLD. An adjudication committee consisting of two board‐certified psychiatrists and a clinical neuropsychologist determined ground‐truth labels for patient diagnostic results of CI (N = 145) or No CI (N = 150). Patients underwent T1‐weighted scans, which were segmented with Freesurfer 6.0.1 (freesurfer.net) packaged with fMRIPrep 1.5.8 (fmriprep.org) using the Desikan‐Killiany atlas. Cortical thickness and volumetric data were obtained along with demographic parameters. Freesurfer data was then regressed against estimated total intracranial volume. 11 classifications algorithms implemented in R were trained and tested on 209 and 88 patients, respectively, to classify if a patient has CI. All continuous‐valued data were normalized with respect to the training set mean and standard deviation. Result Table 1 shows results from multiple classifications algorithms, including high sensitivity, specificity, and accuracy from adaptive boosting algorithms. The latter algorithms iteratively optimize a set of binary classification tree algorithms (Freund et Schapire., 1996). Conclusion Volumetric structural MRI biomarkers trained with machine learning classifiers can adequately categorize depressed older adults with CI which could assist the adjudication process. Future work incorporating structural MRI biomarkers with gold standard clinical data may improve earlier detection of cognitive impairment stemming from a neurodegenerative etiology. References : 1. Tsuno, N., & Homma, A. (2009). What is the association between depression and Alzheimer’s disease?. Expert review of neurotherapeutics, 9(11), 1667‐1676. 2. Freund, Y., & Schapire, R. E. (1996, July). Experiments with a new boosting algorithm. In icml (Vol. 96, pp. 148‐156).
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 | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| 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".