Assessment of the head-twitch response

The head-twitch response (HTR) was assessed using a head-mounted neodymium magnet and a magnetometer detection coil⁷⁸. The mice were allowed to recover from the magnet implantation surgeries for at least 1 week prior to behavioral testing. Mice were tested in multiple HTR experiments, with at least 7 days between studies to avoid carryover effects. HTR experiments were conducted in a well-lit room, and the mice were allowed to habituate to the room for at least 1 h prior to testing. Mice were tested in a 12.5-cm diameter glass cylinder surrounded by a magnetometer coil. Coil voltage was low-pass filtered (2 kHz), amplified, and digitized (20-kHz sampling rate) using a Powerlab (model /8SP or 8/35) with LabChart software (ADInstruments, Colorado Springs, CO, USA). Head twitches were identified in the recordings off-line by their waveform characteristics⁹⁴ or using artificial intelligence⁹⁵. HTR counts were analyzed using one-way ANOVAs or one-way Welch ANOVAs (in cases where groups showed unequal variances). Tukey’s test or Dunnett’s T3 multiple comparisons test was used for post hoc comparisons. Significance was demonstrated by surpassing an α level of 0.05. Median effective doses (ED₅₀ values) and 95% confidence intervals for dose-response experiments were calculated by nonlinear regression (Prism 9.02, GraphPad Software, San Diego, CA, USA).