Gong, Cihun-SiyongCihun-SiyongGongHuang, Zi-HaoZi-HaoHuangLiu, Hao-LiHao-LiLiu2026-04-162026-04-162026-02-02https://www.scopus.com/record/display.uri?eid=2-s2.0-105029349194&origin=resultslisthttps://scholars.lib.ntu.edu.tw/handle/123456789/737226This letter presents an innovative noncontact real-time tracking and focusing treatment system using ultrasound echoes, capable of precise head tracking and acoustic focusing on freely moving objects. The system combines ultrasound echo signals with machine learning to achieve noncontact tracking and stimulation, overcoming traditional limitations that affect natural behavior. It predicts the object's position and adjusts the frequency and pulse width of focused acoustic waves in real time. In experiments, the system achieved an average head tracking error of 0.96 ± 0.61 cm, with sound pressure levels ranging from 69-82 dB and peak intensities of up to 430.27 mW/cm2. The findings indicate that the system may induce neurophysiological responses and, in theory, facilitate neural regulation through acoustic radiation force, thus providing a promising avenue for the advancement of neuro-regulation technology.falseairborne ultrasoundfreely moving target trackinglow-frequency modulationmachine learning coordinate predictionreal-time focus controlSensor systemsultrasonic phased arrayjournal article10.1109/lsens.2026.36603762-s2.0-105029349194