Pixel1.gif (51 bytes)
Pixel1.gif (51 bytes)
Pixel1.gif (51 bytes) Main Page Pixel1.gif (51 bytes)
About DSP Laboratory
Pixel.gif (52 bytes)
Contact Us
Pixel.gif (52 bytes)
Go to FIU's Homepage


 Pixel1.gif (51 bytes)


Curve.gif (104 bytes) Pixel1.gif (51 bytes)

STL: A Spatio-Temporal Charcterization of Focal interictal Events

Pixel1.gif (51 bytes)

"STL: A Spatio-Temporal Charcterization of Focal interictal Events", (1993)
Barreto A.B., Principe J.C. and Reid S.A.

Recent advances in neuroimaging, monitoring and improved neurosurgery have made possible the surgical treatment of some otherwise intractable cases of focal epilepsy. An estimated 200,000 people in the United States have frequent seizures that are not controlled. It has also been observed that 7% to 10% of these suffer from partial simple epilepsy (Crandall et al. 1990), and, as such, are potential candidates for surgical therapy. A number of eligible epileptic patients have experienced significant improvement in their lifestyles after the pathologic areas of their cerebral cortices have been surgically removed or 'resected'.

Two trends have developed in relation to the extent of the resections performed in the surgical procedures for focal epilepsy. The more traditional 'en bloc' approach sets the boundaries for the resection according to anatomical landmarks. On the other hand, a second school of thought is in favor of 'tailoring' the resection according to functional mapping and electrographic findings obtained intraoperatively, through electrocorticography (ECoG). The objective of this second approach is 'to provide the largest resection of epileptogenic cortex that can be safely done without jeopardizing critical neurologic function' (Wyler 1987).

Although, in the opinion of some experts, the epileptic focus could be best identified with the place of onset of a spontaneous seizures, the intraoperative ECoG record may only present interictal events. According to Gloor (1975), these 'still provide direct and otherwise unobtainable evidence of neuronal dysfunction of an epileptiform type and therefore provide valuable diagnostic information if evaluated critically'. Thus, the extent of the 'epileptogenic cortex' will normally be determined by the analysis of the interictal activity represented in the ECoG traces plotted by the EEG machine. These traces are searched for fast transients and, if bipolar chains are used for the recording, phase reversals in derivations involving a common electrode. Then, the areas where the most prominent interictal activity takes place are identified as the epileptic 'foci'. While this kind of information can be retrieved from any appropriate set of ECoG electrodes, an ECoG recording system has been designed (Reid 1989) for this particular application. The system includes a flexible array of 3 by 8 stainless steel electrodes and a movable strip with 5 more electrodes. The rectangular array was designed to be applied over the exposed cerebral cortex, while the strip can be used to sense the electrical activity of inferior or inferomedian structures.

Although the EEGer applies his/her expertise to the analysis of a set of seemingly isolated ECoG traces, this process indeed invokes the mental reconstruction of the potential field configurations that are responsible for the voltage differences displayed in the bipolar traces. According to Gloor (1975), 'The (ECoG or EEG) interpreter must make the mental effort to translate the set of potential-versus-time plots provided by a number of EEG channels into meaningful potential maps. He must therefore be able to visualize the three-dimensional configuration of potential fields of significant electrographic events on the scalp if he is to avoid drawing misleading localizing conclusions'.

In view of the prominent role that the 3-dimensional potential field configurations play in the localization of foci, a computer system could be built to display such 3-dimensional images (or 2- dimensional equivalents, e.g., topographic maps) from the signals recorded with an array like the one described above. However, even if the technical challenge of simultaneous, real-time, data acquisition, processing and display could be overcome, the highly transient nature of the interictal events could complicate the location of their origin from a real-time display. These events, commonly identified as 'spikes' due to their appearance on the paper chart, would produce the characteristic potential field configurations for only 100 milliseconds or less, making it extremely hard for a human observer to determine the focus from then.

Instead, we propose to perform on-line computer analysis of the time-varying potential field, both in space and in time, to determine if a particular time interval involves field configurations that are relevant to the localization of the focus. Thin. only those segments of the stream of incoming data chosen by this analysis will be displayed to the neurosurgeon at a readable speed or in a static fashion.

Our preliminary efforts to reconstruct the potential field perturbation associated with an interictal event, as sensed by the array, indicated that a concave field configuration was characteristic of the event. These observations matched with those noted in numerous experiments with acute induced foci in animal models (Dichter and Spencer 1969; Petsche et al. 1970, 1984, 1987; Rappelsberger et al. 1987; Ayala et al. 1973). They also correlated well with the theoretical potential field predicted for a 'vertical' ('radial') dipole (Nunez 1981) or dipole layer (Gloor 1985). Additionally, the animation of successive reconstructed potential field configurations, in slow motion, allowed us to verify that the concavity of the field evolves according to a time course which essentially follows that of a 'spike', i.e., a sudden, fast increase followed by an abrupt, fast, decrease. This should not be a surprise, since, after all, the bipolar of referential derivation traces, in which this pattern is synonymous with epileptic activity, are just a partial view of the field event that we are considering.

According to this analysis, we conjecture that the concavity of the potential field ('sharpness in space') and the transient character of the phenomenon ('sharpness in time') are the key features of focal interictal activity, as sensed by an ECoG array. Furthermore, we have developed a transformation that when applied to the time-varying potential field sampled in time an space (as represented in the digitized data from the array), yields a new (discrete) time-varying field, which we called the Spatio-Temporal Laplacian (STL) field, In this new field events that are local and transient, such as focal interictal events, will be selectively emphasized.