A machine learning methodology for the selection and classification of spontaneous spinal cord dorsum potentials allows disclosure of structured (non-random) changes in neuronal connectivity induced by nociceptive stimulation

Abstract

Previous studies aimed to disclose the functional organization of the neuronal networks involved in the generation of the spontaneous cord dorsum potentials (CDPs) generated in the lumbosacral spinal segments used predetermined templates to select specific classes of spontaneous CDPs. Since this procedure was time consuming and required continuous supervision, it was limited to the analysis of two specific types of CDPs (negative CDPs and negative positive CDPs), thus excluding potentials that may reflect activation of other neuronal networks of presumed functional relevance. We now present a novel procedure based in machine learning that allows the efficient and unbiased selection of a variety of spontaneous CDPs with different shapes and amplitudes. The reliability and performance of the present method is evaluated by analyzing the effects on the probabilities of generation of different classes of spontaneous CDPs induced by the intradermic injection of small amounts of capsaicin in the anesthetized cat, a procedure known to induce a state of central sensitization leading to allodynia and hyperalgesia. The results obtained with the selection method presently described allowed detection of spontaneous CDPs with specific shapes and amplitudes that are assumed to represent the activation of functionally coupled sets of dorsal horn neurones that acquire different, structured configurations in response to nociceptive stimuli. These changes are considered as responses tending to adequate transmission of sensory information to specific functional requirements as part of homeostatic adjustments.

Supraspinal Shaping of Adaptive Transitions in the State of Functional Connectivity Between Segmentally Distributed Dorsal Horn Neuronal Populations in Response to Nociception and Antinociception

“…We assumed that some clues could be obtained by examining the changes on the probabilities of occurrence of the different classes of ongoing CDPs induced by nociception as well as during the antinociception produced by systemic lidocaine. To this end, we used machine learning (see section “Materials and Methods” and specially Martin et al, 2015) to identify and select the ongoing CDPs recorded in each segment according to their shape and amplitude and examine how capsaicin, lidocaine, and spinalization affected the fractional probabilities of occurrence of each class of CDPs.…”

“…We have previously shown (Martin et al, 2015) that the dictionaries made with the CDPs extracted from control recordings made in different experiments were relatively stable during prolonged time periods (30–60 min). We also found that in a given segment, each time set had a specific dynamical behavior that was changed by the intradermic injection of capsaicin as well as by spinalization.…”

“…To extract the CDP sequences we used the Machine Learning method described in a previous study (Martin et al, 2015). To this end, the whole procedure was divided into several steps that included: (a) the extraction of the CDPs from the raw recordings, (b) the classification and discretization of the selected CDPs, and (c) the analysis of their behavior at different levels of granularity in the spatial and temporal domain.…”

“…To approach the first two questions we utilized Machine Learning procedures that use similarity criteria for the automatic selection and classification of the ongoing CDPs according to their shape and amplitude (see Béjar et al, 2015; Martin et al, 2015). We used this procedure to build dictionaries with CDPs selected under basal (control) conditions to estimate the changes produced during nociception and antinociception on the probabilities of occurrence of the different classes of CDPs, a task that would be otherwise difficult to achieve using predetermined template selection methods such as those employed in previous studies (see Chávez et al, 2012; Contreras-Hernández et al, 2015).…”

Supraspinal modulation of neuronal synchronization by nociceptive stimulation induces an enduring reorganization of dorsal horn neuronal connectivity

“…To this end, we have developed a machine learning procedure for the automatic selection of ongoing CDPs according to their shape and amplitude (Martín et al 2015). With this method the CDPs recorded in a particular experiment during different procedures could be reliably separated in different classes.…”