Key result
Computer-vision hypoxia detectors require ~56 hospital clusters to detect a 5-point sensitivity gain.
Why the study?
Irregular sampling in intrapartum cardiotocography produces non-uniform intervals that violate the fixed-window assumptions of most computer-vision classifiers for fetal hypoxia detection.
Sampling irregularity in intrapartum cardiotocography substantially inflates the number of hospital clusters (to 56) required to adequately power clinical evaluations of computer-vision fetal hypoxia detectors.
May undermine reliability of fixed-window hypoxia classifiers with irregular CTG; leaves open whether adaptive models improve intrapartum detection.
Cardiotocography remains the most widely deployed intrapartum fetal surveillance modality in referral hospitals, yet its sampling is irregular: electrode repositioning, maternal movement and signal loss produce non-uniform intervals that violate the fixed-window assumptions of most computer-vision classifiers for hypoxia detection. the present analysis formalises the resulting temporal-sensitivity problem and evaluates, through a replication-based design analysis, how the before-after-control-impact estimator behaves when the unit of analysis is the hospital and the outcome is a sensitivity proportion. No observational data are used or claimed. Under a declared variance structure with between-cluster SD = 6.0, residual SD = 4.5 and intraclass correlation = 0.64, power was estimated across 2000 replications per design cell for cluster counts of 8, 16, 24 and 32 and effects of 3, 5 and 8 percentage points. The analysis indicates that 56 clusters are required to detect a 5-percentage-point sensitivity gain at 80% power, a requirement that exceeds the referral-hospital networks typically available for such evaluations. The contribution is a design-theoretic account of why sampling irregularity in cardiotocography inflates the cluster requirement for computer-vision hypoxia detectors, and a set of propositions for temporally aware evaluation.
No takes yet. Share an insight, caveat, or question.
Shigwedha et al. (2026) studied Intrapartum fetal hypoxia. Computer-vision classifiers for cardiotocography hypoxia detection under sampling irregularity was evaluated on Cluster requirement to detect a 5-percentage-point sensitivity gain at 80% power. Computer-vision hypoxia detectors under cardiotocography sampling irregularity require 56 hospital clusters to detect a 5-percentage-point sensitivity gain at 80% power (ICC=0.64).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: