MRI edema grades improve retrospective neural-network prognosis for Guillain–Barré syndrome
Source check: Source checked. Not independent scientific confirmation.
Finding: A two-institution retrospective study used a shallow neural network to estimate residual symptoms six months after discharge in 354 Guillain–Barré syndrome patients. Adding radiologist-assigned lumbosacral edema grades to clinical variables increased the source-reported external-cohort AUC from 0.84 to 0.97. The model includes a muscle-strength score measured at discharge, so it is not an admission-only forecast. External testing used 118 patients from a second institution; treatment benefit and prospective deployment were not tested.
Technology: Shallow neural network
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- A shallow neural network learned six-month outcome classification from clinical factors, with and without radiologist-assigned MRI edema grades; comparison isolates the added predictive information in the MRI grade.
- Test and result
- Retrospective development on 236 patients and external testing on 118; source reports external AUC 0.84 for clinical variables versus 0.97 with MRI grading. MRI reader agreement was assessed separately.
- Limitation / still unknown
- The predictor includes discharge muscle strength and does not establish early admission-only utility. Retrospective selection and source reporting inconsistencies limit interpretation; no causal or treatment benefit is inferred.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://doi.org/10.1002/acn3.51984 · https://pmc.ncbi.nlm.nih.gov/articles/PMC10963293/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 21 Nov 2026. Review scope: Exact primary article-front DOI/title/PMCID and date; full relevant model Methods, Results and limitations; fresh PubMed publication types and correction/retraction relationship metadata..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.