Cortical Erosion in Alzheimer's Disease: Testing activity-dependent degeneration with CYTO7 and EEG/MEG
Giulio Ruffini
Cortical dysfunction in Alzheimer's disease may advance through cumulative activity-dependent damage as well as molecular pathology. We propose that sustained excitatory activity beyond local inhibitory and metabolic capacity progressively erodes circuit integrity. Cortical architecture may determine both susceptibility to this process and the connections that distribute afferent drive. CYTO7 provides an anatomy-defined representation of cortical type, while the Structural Model predicts aspects of connection presence, strength and laminar organisation. Together they motivate a test of whether vulnerable temporal mesocortex and its connected territories exhibit a reproducible sequence of electrophysiological change. The primary hypothesis is activity-dependent network erosion; tau transport is a competing or interacting mechanism rather than the definition of progression. We distinguish hyperexcitability, ongoing activity, synchrony and irreversible damage, and use laminar neural mass modelling to relate candidate circuit changes to EEG/MEG observables. The empirical programme combines longitudinal and cross-sectional recordings, cortical-type profiles, connectivity dynamics and, where available, MRI and molecular imaging. It compares local vulnerability, network load, tau exposure and coupled models using independent anatomical priors and held-out outcomes. The intended result is a falsifiable account of regional progression and an interpretable prognostic phenotype, with claims limited by source resolution and the identifiability of excitation and tissue damage from electrophysiology.
A research proposal arguing that Alzheimer's disease might progress the way an overworked machine wears out — not just from spreading toxic proteins, but from circuits repeatedly firing beyond what their inhibition and metabolism can sustain, with the resulting damage rippling outward through the brain's wiring.
The starting puzzle is why some cortical regions (notably temporal mesocortex, an evolutionarily older, thinner-layered tissue near the hippocampus) go bad early in Alzheimer's, and why nearby connected regions follow. The standard story is that misfolded tau protein spreads region to region. This paper proposes a complementary, testable alternative: sustained excitatory activity that outpaces a region's inhibitory and metabolic capacity slowly erodes its circuits, and once damaged, that region's output can push excess load onto whatever it's connected to — a kind of activity-driven cascade rather than a molecular one.
Two existing frameworks do the heavy lifting here. CYTO7 is a cortical atlas that classifies cortex by structural type (how differentiated its layers are, how myelinated) independent of disease, giving a way to say which regions are architecturally fragile before you ever look at pathology. The Structural Model (from Barbas and García-Cabezas' work) predicts, from that same architecture, which regions connect to which, how strong those connections are, and even which cortical layers they hit. Together they sketch both a vulnerability map and a wiring diagram for how bad activity might get redistributed.
The paper is careful to keep four things conceptually separate that EEG/MEG tend to blur together: being prone to overexcite, actually overexciting, becoming abnormally synchronized, and permanently losing circuit integrity. It uses "laminar neural mass models" (biophysical simulations of cortical layers) to work out what each of these would actually look like in scalp recordings, since a change in EEG power alone can't tell you which is happening.
Practically, this is a plan, not a finished study — no pat
- WP ID
- WP0227
- Lifecycle
- ongoing
- Visibility
- internal
- Access level
- open
- Embargo until
- —
- Priority
- —
- Collab
- closed
- Venue
- —
- DOI
- —
- Deadline
- —
- Owner
- giulio.ruffini@bcom.one
- Source
- drive_legacy
- Repo path
- WP0227
- 0.1.0 (draft) · auto-run-placeholder