Cerebellar and Spinal Direct Current Stimulation in Children: Computational Modeling of the Induced Electric Field

Abstract

Recent studies have shown that the specific application of transcranial direct current stimulation (tDCS) over the cerebellum can modulate cerebellar activity. In parallel, transcutaneous spinal DC stimulation (tsDCS) was found to be able to modulate conduction along the spinal cord and spinal cord functions. Of particular interest is the possible use of these techniques in pediatric age, since many pathologies and injuries, which affect the cerebellar cortex as well as spinal cord circuits, are diffuse in adults as well as in children. Up to now, experimental studies of cerebellar and spinal DC stimulation on children are completely missing and therefore there is a lack of information about the safety of this technique as well as the appropriate dose to be used during the treatment. Therefore, the knowledge of electric quantities induced into the cerebellum and over the spinal cord during cerebellar tDCS and tsDCS, respectively, is required. This work attempts to address this issue by estimating through computational techniques, the electric field distributions induced in the target tissues during the two stimulation techniques applied to different models of children of various ages and gender. In detail, we used four voxel child models, aged between 5- and 8-years. Results revealed that, despite inter-individual differences, the cerebellum is the structure mainly involved by cerebellar tDCS, whereas the electric field generated by tsDCS can reach the spinal cord also in children. Moreover, it was found that there is a considerable spread toward the anterior area of the cerebellum and the brainstem region for cerebellar tDCS and in the spinal nerve for spinal direct current stimulation. Our study therefore predicts that the electric field spreads in complex patterns that strongly depend on individual anatomy, thus giving further insight into safety issues and informing data for pediatric investigations of these stimulation techniques.

Keywords: children; computational modeling; ctDCS; high-resolution human models; neuromodulation; tsDCS.

Figures

Figure 1

Segmentation masks for (from left to right) “Roberta”, “Thelonious”, “Eartha” and “Dizzy”. Lateral view of cerebellum (orange), spinal cord (red) and spinal nerves (green) with vertebrae and cerebral tissues in transparency. Black boxes distinguish the four levels of the vertebrae (cervical, thoracic, lumbar and sacral).

Figure 2

Segmentation masks of Eartha’s tissues of interest close to the cerebellum and the spinal cord.

Figure 3

Electrode positioning over model “Eartha”.

Figure 4

Axial section across the cerebellum (2nd column) of the E amplitude distribution for Roberta. The first column on the left shows the plane where the section was calculated. The third column shows the sagittal view of the E amplitude distribution over the cerebellar surface. The colored scale on the right is normalized with respect to the maximum of the E amplitude in the cerebellum.

Figure 5

Descriptive statistic of E amplitude over cerebellum and close brain tissues of the four children models. The boxes indicate the interquartile range (25th–75th), red point the median (or 50th) value and the whiskers the minimum and maximum (or 99th) values.

Figure 6

50th percentile of the E amplitude in the cerebellum (top row), V50 (left) and V70 (right) percentages trends across the four models in comparison with the trends of three anthropometric quantities (i.e., maximum cerebellar antero-posterior length, cerebro spinal fluid (CSF) volume around cerebellum and maximum skull thickness in the occipital bone). The stars identify the anthropometric quantities showing a similar trend to the respective electric field quantity.

Figure 7

Sagittal section across the spinal cord of the E amplitude distribution (2nd column) for Dizzy. The first column on the left shows the plane where the section was calculated. The third column shows a view of the E amplitude distribution over the spinal cord and nerve surface. The color scale on the right is normalized with respect to the maximum of E amplitude in the spinal cord.

Figure 8

Descriptive statistic of E amplitude distribution over the spinal cord at different spine levels, across the four models.

Figure 9

Descriptive statistic of E amplitude distribution over the spinal nerves in different spinal nerve segments, across the three models whose nerves are represented (Eartha not shown).

Figure 10

Trend of the 50th percentile of the E amplitude distribution in the spinal cord (left), and of the CV over transverse section (right) across the four models in comparison with the trend of three anthropometric quantities at thoracic (top) and lumbar (down) level (i.e., CSF volume, spinal cord volume and spinal cord length). The stars identify the anthropometric quantity that presents a similar trend to the respective electric field quantity.