Summary information of the six preselected methods
| Approach | i | ii | iii | iv | v | vi |
|---|---|---|---|---|---|---|
| Detailed description and usage information | Yes | Yes | Yes | Yes | Yes | Yes |
| Parametrization and adaptivity | Internal adaptivity is due to the nature of the algorithm | The ISI distribution reflects the data specifics | The ISIN distribution reflects the data specifics | Automatic adaptive firing rate threshold is present. | A likely good fit to the data may be achieved with parameter tuning | Four main parameters are available to adjust the detection |
| Code obtained from the author | Yes | Yes | Yes | Yes | No | No |
| Code description | Yes | Yes | Yes | Yes | No | No |
| Verdict for acquired data | Some tendency to break NBs into shorter components | Good performance but NB borders appeared to be widened; NB merging due to tonic spiking in, e.g., one channel | Sufficient adaptivity for the test data; NB borders assigned relatively well | The method detected some short NBs in addition to evident synchronous patterns | Not available | Not available |
Visualizations and a quantitative comparison of the NB detection results for approaches i–iv with available codes are shown in Extended Data Table 1-1.