Quality Metrics module¶
Quality metrics allows to quantitatively assess to goodness of a spike sorting output.
qualitymetrics sub-module includes functions to compute a large variety of available metrics.
All of the metrics currently implemented in spikeInterface are per unit (pairwise metrics do appear in the literature).
Each metric aims to identify some quality of the unit. Contamination metrics (also sometimes called ‘false positive’ or ‘type I’ metrics) aim to identify the amount of noise present in the unit. Examples include: ISI violations, sliding refractory period violations, SNR, NN-hit rate. Completeness metrics (or ‘false negative’/’type II’ metrics) aim to identify whether any of the spiking activity is missing from a unit. Examples include: presence ratio, amplitude cutoff, NN-miss rate. Drift metrics aim to identify changes in waveforms which occur when spike sorters fail to successfully track neurons in the case of electrode drift.
Some metrics make use of principal component analysis (PCA) to reduce the dimensionality of computations. Various approaches to computing the principal components are possible, and choice should be carefully considered in relation to the recording equipment used. The following metrics make use of PCA: isolation distance, L-ratio, D-prime, Silhouette score and NN-metrics. By contrast, the following metrics are based on spike times only: firing rate, ISI violations, presence ratio. And amplitude cutoff and SNR are based on spike times as well as waveforms.
For more details about each metric and it’s availability and use within SpikeInterface, see the individual pages for each metrics.
- Amplitude median (not yet implemented)
- Amplitude cutoff (
- Cumulative drift (not yet implemented)
- D-prime (
- Firing rate (
- Inter-spike-interval (ISI) violations (
- Isolation distance (
- L-ratio (
- Maximum drift (not yet implemented)
- Nearest Neighbor Metrics (
- Noise cutoff (not yet implemented)
- Presence ratio (
- Silhouette score (not currently implemented)
- Sliding refractory period violations (not yet implemented)
- Signal-to-noise ratio (
This code snippet shows how to compute quality metrics (with or without principal components) in SpikeInterface:
we = WaveformExtractor.load_from_folder(...) # start from a waveform extractor # without PC metrics = compute_quality_metrics(we, metric_names=['snr']) assert 'snr' in metrics.columns # with PCs pca = WaveformPrincipalComponent(we) pca.set_params(n_components=5, mode='by_channel_local') pca.run() metrics = compute_quality_metrics(we) assert 'isolation_distance' in metrics.columns
For more information about quality metrics, check out this excellent documentation from the Allen Institute.