PANDIA: Infant Pain Assessment Performance Comparison Across Four Datasets

2,847 infant records from four datasets were used to evaluate the PANDIA multimodal AI system for pain assessment in neonates aged 0–3 months. The dataset likely contains performance metrics from the comparison, which achieved 87.3% accuracy and a 12.4% improvement over baselines. The dataset was published by Oussama El Othmani on figshare in May 2026.

Use Cases

• Benchmark multimodal AI models for infant pain assessment based on the described accuracy metrics.
• Evaluate the impact of personalized meta-learning adaptation on clinical assessment performance.
• Study the trade-offs between model interpretability (via concept bottlenecks) and predictive accuracy.
• Analyze federated learning performance for privacy-preserving multi-site medical data collaboration.

Strengths

• Performance metrics are derived from a substantial sample of 2,847 infants.
• Results are validated across four independent datasets and an out-of-distribution test set.
• The associated AI system is fully publicly available with code, models, and pipelines.
• The dataset is licensed under CC-BY-4.0 for open reuse.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The description notes dataset heterogeneity across collection sites as a key limitation.

Provenance

Source

figshare

Collection Method

Likely compiled from experimental results evaluating the PANDIA AI system.

Freshness

Last updated 2026-05-26 17:32:06; freshness should be verified.