1. The technique of intracortical microstimulation (ICMS), supplemented by single-neuron recording, was used to carry out an extensive mapping of the face primary motor cortex. The ICMS study involved a total of 969 microelectrode penetrations carried out in 10 unanesthetized monkeys (Macaca fascicularis). 2. Monitoring of ICMS-evoked movements and associated electromyographic (EMG) activity revealed a general pattern of motor cortical organization. This was characterized by a representation of the facial musculature, which partially enclosed and overlapped the rostral, medial, and caudal borders of the more laterally located cortical regions representing the jaw and tongue musculatures. Responses were evoked at ICMS thresholds as low as 1 microA, and the latency of the suprathreshold EMG responses ranged from 10 to 45 ms. 3. Although contralateral movements predominated, a representation of ipsilateral movements was found, which was much more extensive than previously reported and which was intermingled with the contralateral representations in the anterior face motor cortex. 4. In examining the fine organizational pattern of the representations, we found clear evidence for multiple representation of a particular muscle, thus supporting other investigations of the motor cortex, which indicate that multiple, yet discrete, efferent microzones represent an essential organizational principle of the motor cortex. 5. The close interrelationship of the representations of all three muscle groups, as well as the presence of a considerable ipsilateral representation, may allow for the necessary integration of unilateral or bilateral activities of the numerous face, jaw, and tongue muscles, which is a feature of many of the movement patterns in which these various muscles participate. 6. In six of these same animals, plus an additional two animals, single-neuron recordings were made in the motor and adjacent sensory cortices in the anesthetized state. These neurons were electrophysiologically identified as corticobulbar projection neurons or as nonprojection neurons responsive to superficial or deep orofacial afferent inputs. The rostral, medial, lateral, and caudal borders of the face motor cortex were delineated with greater definition by ICMS and these electrophysiological procedures than by cytoarchitectonic features alone. We noted that there was an approximate fit in area 4 between the extent of projection neurons and field potentials anti-dromically evoked from the brain stem and the extent of positive ICMS sites.(ABSTRACT TRUNCATED AT 400 WORDS)
Dual Semi-Supervised Learning for Facial Action Unit Recognition
