Decoding Kinematics Using Task-Independent Movement-Phase-Specific Encoding Models

Abstract

Neural decoders of kinematic variables have largely relied on task-dependent (TD) encoding models of the neural activity. TD decoders, though, require prior knowledge of the tasks, which may be unavailable, lack scalability as the number of tasks grows, and require a large number of trials per task to reduce the effects of neuronal variability. The execution of movements involves a sequence of phases (e.g., idle, planning, and so on) whose progression contributes to the neuronal variability. We hypothesize that information about the movement phase facilitates the decoding of kinematics and compensates for the lack of prior knowledge about the task. We test this hypothesis by designing a task-independent movement-phase-specific (TI-MPS) decoding algorithm. The algorithm assumes that movements proceed through a consistent sequence of phases regardless of the specific task, and it builds one model per phase by combining data from different tasks. Phase transitions are detected online from neural data and, for each phase, a specific encoding model is used. The TI-MPS algorithm was tested on single-unit recordings from 437 neurons in the dorsal and ventral pre-motor cortices from two nonhuman primates performing 3-D multi-object reach-to-grasp tasks. The TI-MPS decoder accurately decoded kinematics from tasks it was not trained for and outperformed TD approaches (one-way ANOVA with Tukey's post-hoc test and -value <0.05). Results indicate that a TI paradigm with MPS models may help decoding kinematics when prior information about the task is unavailable and pave the way toward clinically viable prosthetics.

Decoding of standard and non-standard visuomotor associations from parietal cortex

“…In addition, they implemented an extended model to decode multiple states, one for each reaching goal. In another study a four states (additional holding state) HMM was used to detect hidden neural states and so to develop a task independent decoder [57]. Here we used a simpler, but equally informative, Bayesian decoder to obtain posterior probabilities of free, delay and movement states.…”

“…A large amount of evidence has already reported that single PPC neurons can encode both spatial (sensory) and non-spatial (cognitive) information [52,[57][58][59][60]. For example, attention toward a specific spatial location or toward non-spatial visual features modulate lateral intraparietal neurons [50,60,61], parietal reach region encodes both the target location and the movement intention [58,59].…”

Modulations in Oscillatory Activity of Globus Pallidus Internus Neurons During a Directed Hand Movement Task—A Primary Mechanism for Motor Planning

“…Point process methods have been used to analyze the spike train activity for a broad range of neural systems (Sarma et al, 2010, 2012; Saxena et al, 2010, 2012; Santaniello et al, 2012; Sumsky et al, 2017; Sumsky and Santaniello, 2018). A neural spike train can be treated as a stochastic series of random binary events (i.e., the spike times) continuously occurring in time, otherwise known as a point process (Truccolo et al, 2005; Coleman and Sarma, 2010; Sarma et al, 2010).…”

Decoding spatial locations from primate lateral prefrontal cortex neural activity during virtual navigation

“…Most BCI implementations decode the movement intentions from motor cortex neuronal activity to control assistive devices (Collinger et al 2013, Shenoy et al 2013. Recently, studies have demonstrated the possibility to better guide iBCI effectors using discrete movement intentions (Fan et al 2014), states (Kao et al 2017, Sumsky et al 2017, transitions (Kang et al 2015), and goals (Andersen et al 2019) decoded from other cortical areas of the brain. In line with this strategy, our results show that spatial locations within a physical or virtual environment can be deciphered from the LPFC intracortical activity, possibly to assist in the control of intelligent wheelchairs.…”