Predictive Warning of Nociceptive Temperature during Prosthetic Hand Prehension

Grasping hot objects may be harmful, and a warning of nociceptive temperature should be provided to prevent prosthetic hands from damage. This study developed a fast algorithm to predict the steady state temperature of grasped objects based on a thermoresistant sensor, and verified a bi-state sensory encoding strategy to inform either contact force or nociceptive temperature at prosthetic finger. Below the threshold, a buzz percept was used to encode fingertip force, and above the threshold, a tingling pain sensation was induced to warn harmful temperature. This sensory encoding mechanism was tested in one able-bodied subject with a psychophysical experimental paradigm. Results indicated that fast identification of the steady state temperature could be achieved in 0.75±0.00 s with an error of 8.3±6.0%. The subject was able to recognize whether the object temperature was above or below the threshold of nociceptive temperature with an average success rate of 98.2% and 97.0%, respectively. The action to release the tube could be executed in 0.80±0.32 s. The reaction time of successful identification and execution was 1.55 ± 0.32 s. The probability that a release reaction was triggered at the threshold of 60.0°C was calculated with tube temperatures varying from 30.0°C to 80.0°C. The 50% percentile point on the fitted stimulus-response curve corresponded to a temperature of 59.5°C, nearly identical to the internally set threshold. The psychophysical behavior test verified the sensory function to recognize and notify unsafe temperatures in real time.Clinical Relevance-The method of identifying and alerting nociceptive temperatures may restore the sensory ability of amputees to avoid potential damage of grasping hot objects with a prosthetic hand in activities of daily life.