Assessing the utility of SISCOM and PISCOM in mapping the epileptogenic zone for epilepsy surgery: A comparative outcome-based study

Intro

Ictal SPECT imaging stands out among non-invasive tools as the primary presurgical method for spatially defining the seizure network. Advanced image analysis methods like SISCOM and PISCOM improve the interpretation of SPECT images. However, research on their comparative utility, significance of their overlapping areas, and optimal z-scores for mapping/localizing the seizure onset zone is limited. Therefore, this study concentrates on a detailed analysis of these images in mapping of ictal networks and localization of the seizure onset zone.

Methodology

This study analyzed SOZ localization in 37 patients using SISCOM and PISCOM, corroborated by postoperative MRI and outcome. We aimed to determine optimal z-scores for both methods and investigated the influence of the SPECT injection time-seizure duration ratio on interpretation of these maps in SOZ localization accuracy. We also created a ratio image representing the overlap between SISCOM and PISCOM maps, using Z-scores greater than 1, 1.5, and 2 for both maps.

Results

Of 21 Engel-1 patients, in 16 (76 %) the SISCOM and PISCOM perfusion (large, high intensity) lateralized to the resection side, in 3 (14 %) patients to the contralateral side, and in 2 (10 %) showed bilateral activation. In Engel 1 patients, hyperperfusion on the PISCOM mapped the resection zone in 90 % of cases, compared to 80 % with the SISCOM map. The optimal z-scores were 1.9 for SISCOM, yielding a sensitivity of 88.5 % and specificity of 94.8 % for mapping the SOZ. For PISCOM, the optimal z-score was 2.2, achieving a sensitivity of 87.4 % and specificity of 93.3 %. The ratio map, created from SISCOM and PISCOM overlap (z-scores > 2), was useful for delineating the SOZ and aligned better with the resection cavity than SISCOM and PISCOM maps. The analysis showed no significant median differences in injection-time-to-seizure-duration ratios between seizure-free and non-seizure-free patients (p = 0.8). We found a difference in the median injection-time-to-seizure-duration ratio between cases where SISCOM/PISCOM maps were concordant with the resection cavity in seizure-free patients, at 0.09 (0.06, 0.28), and discordant cases, at 0.41 (0.17, 0.61).

Conclusion

This study determined the optimal Z-score for mapping the SOZ and evaluated the use of ratio image in SOZ localization. Both SISCOM and PISCOM demonstrated their utility in identifying the SOZ in epilepsy surgery patients, with PISCOM emerging as a promising alternative to SISCOM, potentially eliminating the need for interictal SPECT.