NeuroMatch Score

Evaluate your event detection algorithms with real rigor.

NeuroMatch Score is a clean, Python-based toolkit that compares your detected events to ground truth using time- and amplitude-matched scoring, per-bin F1 analysis, and redundancy-aware evaluation metrics.

Built for neuroscientists, electrophysiologists, and data scientists working with real biological signals.

Zero fluff. Just meaningful insight and clean visuals.

💡 Features:

• Global event matching with cost matrix (Hungarian algorithm)
• Customizable time window, amplitude weighting
• Per-bin F1 score analysis across amplitude bins
• Redundancy penalties to discourage detector spam
• Clean CSV output and Matplotlib plots

💾 Perfect for:

• Comparing automated detectors vs. ground truth
• Benchmarking ML or template-based spike/IPSC/EPSC finders
• Saving hours of manual validation