From thought to action: The brain-machine interface in posterior parietal cortex

Abstract

A dramatic example of translational monkey research is the development of neural prosthetics for assisting paralyzed patients. A neuroprosthesis consists of implanted electrodes that can record the intended movement of a paralyzed part of the body, a computer algorithm that decodes the intended movement, and an assistive device such as a robot limb or computer that is controlled by these intended movement signals. This type of neuroprosthetic system is also referred to as a brain-machine interface (BMI) since it interfaces the brain with an external machine. In this review, we will concentrate on BMIs in which microelectrode recording arrays are implanted in the posterior parietal cortex (PPC), a high-level cortical area in both humans and monkeys that represents intentions to move. This review will first discuss the basic science research performed in healthy monkeys that established PPC as a good source of intention signals. Next, it will describe the first PPC implants in human patients with tetraplegia from spinal cord injury. From these patients the goals of movements could be quickly decoded, and the rich number of action variables found in PPC indicates that it is an appropriate BMI site for a very wide range of neuroprosthetic applications. We will discuss research on learning to use BMIs in monkeys and humans and the advances that are still needed, requiring both monkey and human research to enable BMIs to be readily available in the clinic.

Keywords: brain–machine interface; intention; monkey; posterior parietal cortex; tetraplegia.

Fig. 1.

Map of intentions within the PPC. Within the intraparietal sulcus are regions selective for intending reaches (the PRR), saccades (the LIP), and grasps (the AIP). Modified from ref. .

Fig. 2.

Examples of different AIP neurons activated by attempted or imagined movements of the left and right hand or speaking left and right. (A) Example neuron coding imagined and attempted movements of the right hand only. (B) Example neuron coding attempted movements of the left hand only. (C) Example neuron coding imagined movements of the right and left hand. (D) Example neuron active to speaking “Left.” Modified from ref. . Copyright (2017), with permission from Elsevier.

Fig. 3.

(A) Example of mixed coding that is random and has no structure. (B) Schematic of the structure of partially mixed selectivity in human AIP. The example is for activity of 3 neurons, in which the shoulder and hand are more separated, and the cognitive strategy and side of the body are more overlapping. Reprinted from ref. . Copyright (2017), with permission from Elsevier.

Fig. 4.

(A) Locations of the front and back contralateral arm sensed by ICMS. (B) Sensations elicited by stimulation. Note that the reported sensations were natural. Modified from ref. , which is licensed under CC BY 4.0.

Fig. 5.

Example monkey neurons from PPC that show intrinsic variable learning. The upper row shows a trained neuron. The left shows that the neuron was active for reaches to stimulus 2 and less so for stimulus 1. The middle shows activity of the same neuron in the BMI pro task in which the animal moves a cursor toward stimulus 1 and 2 locations. The right shows activity of the neuron in the BMI anti task in which the neuron was trained to flip its activity to move the cursor opposite to the direction of stimulus 1 and stimulus 2. The bottom 3 rows show neurons that were recorded simultaneously with the trained neuron, but were not trained. They also flipped their turning, consistent with intrinsic variable learning. Reprinted from ref. . Copyright (2013), with permission from Elsevier.