Fig 1.
Conventional structural MRI features of focal cortical dysplasia in a representative subject (patient 1) for 3 orthogonal slices.
The abnormalities are indicates by arrows; in the T1-weighted images (A) these are cortical thickening and blurring of the gray matter-white matter interface; in the FLAIR-weighted images (B), hyperintensities in the underlying white matter can be observed.
Table 1.
Patient and FCD characteristics, and descriptions of the connectivity profiles.
Aberrant connectivity profiles (compared to controls) are found in most cases; typically the abnormality in local functional connectivity extended beyond the structural lesion. SF: seizure frequency; low signifies one seizure per month, high signifies multiple seizures per week. IISS: inter ictal seizure spread; low signifies confined to one or several ipsilateral electrodes, high signifies including contralateral electrodes. n.d.: none detected; no epileptiform activity found during EEG acquisition. n.a.: not applicable.
Fig 2.
Resting-state fMRI processing pipeline.
The cortical surface is segmented from the T1-weighted scan and the preprocessed fMRI data is mapped to the cortex. The gyral pattern is normalized and maps are derived for distance and seed-based functional connectivity to construct profiles of connectivity as a function of distance. These are investigated to find abnormalities in local functional connectivity for the individual patients.
Fig 3.
Inflated views of the pial surface of 3 representative subjects; the pattern of gyri/sulci is encoded in blue/red.
Alternatively, these gyral patterns may be represented in spherical view. These spherical views can be mapped to a spherical standard space in Freesurfer, which is associated with an anatomical standard space. Data associated with the gyral pattern, such as cortical locations (black markers), are thus mapped to an anatomical standard, which enables homotopic between-subject comparisons.
Fig 4.
For the cortical location indicated by an asterisk (*), the shortest paths (along the cortex) to a number of other cortex vertices were determined and color coded by their path length in A; combining this for all cortex vertices yields the geodesic distance map in B.
Fig 5.
For one exemplary patient (patient 1), the lesion location and its connectivity map are given in A; the inset is the average homotopic connectivity map of the controls.
The geodesic distance map for this location is given in B. By combining the connectivity and distance maps, connectivity profiles as depicted in C can be constructed. Also average homotopic connectivity profiles for the healthy control subjects and for the other patients (i.e. excluding patient 1) are provided in C, showing no significant differences in connectivity profiles between the healthy controls and the other patients (without lesions at the lesion location of patient 1). The extent of the structural lesion on conventional MRI is shaded; connectivity values of the single patient that are aberrant compared to the set of normative control curves are indicated by an asterisk (*). Surface maps represent inflated views; in contrast, connectivity values and distance maps were assessed in native space. Error bars represent 1 standard error.
Fig 6.
Different profiles of aberrant connectivity.
In patient 2, transient hypoconnectivity is found directly beyond the lesion visible on conventional structural MRI (A). In patient 5, on the other hand, transient distal hyperconnectivity is found (B). In patient 10, a dual profile is found of hyperconnectivity within and directly beyond the structural lesion, and hypoconnectivity more distally (C). In a few patients, no aberrant connectivity values were found, such as in patient 4 (D). Error bars represent 1 standard error.