Home › Research › Brain scans and imaging
Information-Theoretic Geometry HCFIM features for data-limited EEG screening of Alzheimer’s disease and Frontotemporal Dementia
Rojas-Pescio H.. Research Square, 2026-10-13
Topic: Brain imaging, connectivity and assessment
Resting-state EEG is a scalable modality for neurophysiological screening, but conventional spectral descriptors capture only part of the signal organization altered in neurodegenerative disease. This study evaluates a reproducible information-theoretic geometry framework based on entropy (H), statistical complexity (C) and Fisher Information Measure (FIM) for data-limited EEG characterization of Alzheimer’s disease (AD) and frontotemporal dementia (FTD). Using the public OpenNeuro resting-state EEG dataset comprising 29 healthy controls, 36 AD patients and 23 FTD patients, preprocessed eyes-closed 19-channel recordings were segmented into epochs, transformed into finite probability distributions, projected into HC and HFIMplanes and HCFIMspace, and aggregated at subject level. Synthetic signal regimes were first used to verify the behavior of the probability-domain geometry. Regional and lateralized cortical schemas were then compared against PSD baselines using leakage-aware subject-level validation. Then, various machine learning classification models were applied to distinguish between healthy and neurologically affected individuals. In the regional schema, the most consistent AD-related geometric displacement was found in temporal alpha rhythms, using Permutation entropy with m = 4, τ = 1 (inter/intra ratio = 0.847; silhouette = 0.141), while in FTD it was found in temporal broadband (inter/intra ratio = 1.140; silhouette = 0.073); both silhouette values indicate modest, overlapping group geometry rather than well-separated clusters. For AD classification, PSD + HCFIM with Extra Randomized Trees achieved ROC AUC = 0.818 and balanced accuracy = 0.803; adding lateralized differences increased ROC AUC to 0.832 and sensitivity to 0.833, with lower specificity. For FTD, regional HCFIM-only features achieved the highest ROC AUC (0.810), while regional PSD + HCFIM achieved the best balanced accuracy (0.740). Lateralization highlighted left temporal broadband displacement in FTD but did not improve classification. These results support HCFIM as an interpretable, reproducible complement to spectral EEG features for preclinical methodological research, while emphasizing the need for independent validation before clinical or device-specific claims.
Key findings
- These results support HCFIM as an interpretable, reproducible complement to spectral EEG features for preclinical methodological research, while emphasizing the need for independent validation before clinical or device-specific claims.
New to this subject? Start with our plain-English guide: Brain scans & tests →
Study details
- Effect sizeAUC = 0.818
- Evidence strengthHigh
How this study scored 16/100
Preprint — not yet peer-reviewed. It scored 32 on the same rubric, but non-peer-reviewed sources are capped at 16, so it shows as 16. Scored zero for study design, sample size, studied in people and randomised / blinded because the record does not state it. We have no data on times cited, citations per year, impact for its field and year, cited by studies here, still being cited, senior author's track record, institution's track record and multi-centre collaboration, so those were left out: it earned 20 of the 62 points we could judge it on, which is 32 out of 100.
| Component | Points | Why |
|---|---|---|
| Journal tier | 2 / 10 | Tier 4 — not a peer-reviewed journal (Research Square) |
| Journal's citation record | 0 / 4 | Not a journal article, so the journal's record does not count |
| Study design | 0 / 12 | No study design recorded — scores zero, not guessed (not stated in the record — scored 0) |
| Strength of evidence | 5 / 5 | Rated “high” in our summary |
| Peer-review validation | 4 / 10 | Preprint — not yet peer-reviewed. Hosted on a preprint server. |
| How recent | 5 / 5 | 2026 — within 1 year(s) of the newest study here |
| Sample size | 0 / 4 | Participant count not stated — scores zero (not stated in the record — scored 0) |
| Effect reported | 3 / 3 | Effect reported with a statistic (interval, p-value, ratio or similar) |
| Studied in people | 0 / 3 | Not stated whether this was studied in people — scores zero (not stated in the record — scored 0) |
| Randomised / blinded | 0 / 2 | No randomisation or blinding stated — scores zero (not stated in the record — scored 0) |
| Times cited | — / 7 | No citation count (not found in OpenAlex) — left out of this score |
| Citations per year | — / 5 | No citation count or no year — left out of this score |
| Impact for its field and year | — / 6 | No field-weighted citation impact (OpenAlex has none yet) — left out of this score |
| Cited by studies here | — / 6 | Not yet checked against this library's citation graph — left out of this score |
| Still being cited | — / 4 | No yearly citation counts — left out of this score |
| Senior author's track record | — / 4 | Senior author's record not available — left out of this score |
| Institution's track record | — / 3 | No institution listed in OpenAlex — left out of this score |
| Multi-centre collaboration | — / 3 | Institutions not listed — left out of this score |
| Indexed in PubMed | 1 / 2 | Not in PubMed, but has a DOI |
| Free to read | 0 / 2 | Behind a paywall — scores zero |
| Points | 20 / 62 | Scored on 12 of 20 parts. We have no data for the rest, so they are left out: 20 of the 62 points we could judge it on is 32 out of 100. |
| Total | 32 / 100 | Capped to 16 because this is not a peer-reviewed article. |
What each of these 20 components measures, and what this score is not →
Read the original
Free copy unknown: no free-copy check has an answer for this record yet. That does not mean it is paywalled.
Cite this study APA, Vancouver, BibTeX · download .bib or .ris
Built from the record alzhub holds for this paper, not looked up. Volume, issue and page numbers are not held in the record. Check the citation against the original before you use it.
APA 7th
Rojas-Pescio, H. (2026). Information-Theoretic Geometry HCFIM features for data-limited EEG screening of Alzheimer’s disease and Frontotemporal Dementia. Research Square. https://doi.org/10.21203/rs.3.rs-10834216/v1
Vancouver
Rojas-Pescio H. Information-Theoretic Geometry HCFIM features for data-limited EEG screening of Alzheimer’s disease and Frontotemporal Dementia. Research Square. 2026. doi:10.21203/rs.3.rs-10834216/v1.
BibTeX
@article{rojaspescio2026-f194ed38,
author = {Rojas-Pescio, H.},
title = {{Information-Theoretic Geometry HCFIM features for data-limited EEG screening of Alzheimer’s disease and Frontotemporal Dementia}},
journal = {Research Square},
year = {2026},
doi = {10.21203/rs.3.rs-10834216/v1},
url = {https://doi.org/10.21203/rs.3.rs-10834216/v1}
}
Related studies in Brain scans and imaging
Also tagged “convolutional”
205 studies in the library carry the topic tag “convolutional”; these are this study's neighbours in a fixed, alphabetical-by-hash tour of them, not the closest in content. All 205 tagged studies.