#1 · ML / AI methods
Joint inference of disease time, progression subtypes, and multimodal trajectories
The dominant thread is subject-level modeling of change over time rather than cross-sectional comparison. A connectome-constrained dynamic model jointly infers each subject's disease time and data-driven progression subtypes from longitudinal Parkinson's morphometry in PPMI, while TractSpLearn applies a shared-manifold framework to DKI tract features for individualized detection of subtle white matter alterations in mild traumatic brain injury. What is new in the first is the joint estimation of time and subtype within one scalable model. The remaining items address adjacent methodological concerns: pyAvalanches standardizes spatiotemporal propagation analysis of EEG neuronal avalanches, and AnnoAudit shows that a permissive marker rule can manufacture a spurious neuronal population in a TBI single-cell atlas. These items share a methodological rather than biological endpoint.
Preprints
Discovering Subtypes of Neurodegenerative Progression with a Scalable Connectome-Constrained Dynamic Model
Connectome-constrained dynamic model jointly infers subject disease time and data-driven progression subtypes from longitudinal Parkinson's morphometry (PPMI),…
pyAvalanches: A Python Package for Analyzing Spatiotemporal Propagation in Neuronal Avalanches
pyAvalanches is an open-source Python package standardizing neuronal avalanche analysis from EEG, including Avalanche Transition Matrices for spatiotemporal…
TractSpLearn: Specialized Shared-Manifold Learning for Individualized Detection of Subtle White Matter Alterations in Mild Traumatic Brain Injury
TractSpLearn is a shared-manifold learning framework using DKI tract features to detect individualized subtle white matter alterations in mild TBI,…
AnnoAudit: a marker-based protocol for auditing single-cell atlas annotations reveals annotation-driven artifacts in a widely used traumatic brain injury resource
AnnoAudit, a marker-based protocol auditing single-cell atlas labels, finds a permissive rule manufactured a spurious neuronal population in a TBI dataset…
Topic history
| Date | Rank |
|---|---|
| #1 | |
| #2 | |
| #36 |