AI and Neurological Rare & Neuroimmune Diseases ── ALS ── 2026-11-04
ALS: treatment, trials and AI
- Abstract
A systematic review (PRISMA 2020) of artificial intelligence and machine learning methods used to construct, augment or adjust the control arm of a clinical trial. Three questions: which methods have been used, in which diseases and from which data sources; what their methodological quality, risk of bias and regulatory credibility are; and which are transferable to amyotrophic lateral sclerosis. Eligibility turns on function rather than terminology: a study is included if a learned model substitutes for or adjusts the trial's control arm, whatever the authors call it. Purely statistical approaches with no learned component are excluded. Each included study is appraised in three layers: PROBAST+AI for the prediction model, a six-item causal validity checklist for the treatment-effect inference, and a descriptive mapping against the FDA seven-step credibility assessment framework. Method families are then scored against eleven ALS-specific requirements. Narrative synthesis; no meta-analysis planned.
Journal IF: unknown (could not be matched)Reference: [No authors listed]. Artificial intelligence methods for synthetic and external control arms in clinical trials: a systematic review with credibility appraisal and translational assessment for amyotrophic lateral sclerosis. Open Science Framework. [Epub ahead of print]. doi:10.17605/osf.io/bfj6c.Checked: Abstract only