FibreCastML: Electrospun Nanofibre Diameter Distribution Predictions

68,538 fibre-diameter measurements from 1,778 studies across 16 biomedical polymers were curated to train machine learning models. The FibreCastML framework predicts full fibre diameter spectra from six standard electrospinning processing parameters. The dataset and associated web application were created by Elisa Roldán and last updated in March 2026.

Use Cases

• Predicting full fibre diameter distributions based on solution concentration, voltage, and flow rate.
• Optimizing electrospinning process parameters for specific biomedical polymers like Polyvinyl alcohol or Nylon-6.
• Generating interpretable insights into parameter influence using SHAP values and feature importance.
• Validating experimental electrospinning results against model-predicted diameter spectra.

Strengths

• Meta-dataset contains 68,538 measurements from 1,778 studies, providing a substantial foundation for model training.
• Models achieved R² > 0.91 for nine specific polymers, indicating high predictive performance.
• Framework validated experimentally with a reported distribution overlap coefficient of 84%.
• Includes six clearly defined input parameters (e.g., concentration, voltage, tip-to-collector distance).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the underlying training data is unknown, which may limit suitability assessment.
• Data may reflect publication bias inherent to the meta-analysis of 1,778 existing studies.

Provenance

Source

figshare, author Elisa Roldán.

Collection Method

Comprehensive meta-dataset curated from 1,778 studies.

Freshness

Last updated 2026-03-17 13:08:27.