Improved sleep spindle detection in rats using the BOSC method: A comparison with the traditional automated approach

Sleep spindles are bursts of 6-20 Hz field activity that occur during non-rapid eye movement sleep. Traditional spindle detection methods rely on manually set amplitude and duration thresholds, but this approach can be vulnerable to false detections and could miss low-amplitude spindles due to their inflexible natures.

The Better OSCillation (BOSC) detection method offers a robust alternative by applying frequency-specific power thresholds calibrated to the aperiodic aspects of the signal itself and also requiring a minimum number of oscillatory cycles. Here we compared the most traditional method of spindle detection against BOSC in a variety of ways. First, we created two synthetic datasets: one with synthetic spindles modelled on previous data and one with single wave pulses that had a period consistent with the spindle frequency band. These events were added to random noise with 1/f power properties as well as true local field potential backgrounds. These datasets were used to demonstrate the performance of each method when presented with events that should be detected (synthetic spindles) and those that should not (single wave pulses).

In every case, BOSC outperformed the traditional method. We then analyzed cortical EEG from rats during natural sleep using both methods to determine their relative effectiveness given real data. BOSC consistently detected more spindle-like activity in all situations and minimized false (or likely false) detections as compared to the traditional method.

These findings validate BOSC as a superior method of spindle detection.