Cavitational capacitive drive: a computationally efficient model for ultrasonic neuromodulation

Objective.Ultrasonic neuromodulation is emerging as a promising non-invasive technique for modulating neuronal activity. Among the proposed mechanisms, the neuronal intramembrane cavitation excitation (NICE) model provides a biophysically grounded description of ultrasound (US)-membrane interactions. However, the computational complexity of the NICE framework results in prolonged simulation times, limiting its applicability to large-scale and multicompartment neuronal models. This study presents a computationally efficient and easy-to-implement approximation of the NICE model termed the cavitational capacitive drive (CCD) model.Approach.The CCD model reproduces the US-induced membrane capacitance oscillations generated by the NICE framework using an analytical formulation parameterized by US frequency and intensity. The model was calibrated against NICE-generated capacitance waveforms and implemented as a distributed membrane mechanism in the NEURON simulation environment. Model performance was evaluated by comparing the passive and active neuronal responses predicted by the CCD and NICE models. The model was tested for the US frequency range from 100 to 1000 kHz, and intensity range from 10 to 2000 mW cm, suitable for continuous wave ultrasonic neuromodulation.Main results.The effective membrane capacitance predicted by the CCD model showed excellent agreement with the NICE model across the investigated stimulation range (). The CCD model accurately reproduced NICE-derived changes in passive membrane properties of a Hodgkin-Huxley neuron and active responses of a cortical regular-spiking neuron. Despite maintaining high accuracy, the CCD model achieved an average computational speed-up of more than 8,500-fold relative to the NICE framework. To demonstrate the capability of our approach, we applied it to multicompartment neuron models, showing that its computational efficiency allows the investigation of US-induced changes in cable properties, synaptic potential propagation, and action-potential conduction.Significance.By replacing the computationally intensive electromechanical calculations of the NICE model with a direct capacitance-based formulation, the CCD model substantially reduces simulation cost while preserving the key neuromodulatory effects predicted by NICE. The proposed framework facilitates the incorporation of intramembrane-cavitation-based ultrasonic neuromodulation into complex neuronal models and provides a practical tool for large-scale computational studies.