Quantifying Similarity Between Perfusion Pattern in Repeat Ictal SPECT Injection
Abstract number :
3.25
Submission category :
5. Neuro Imaging / 5B. Functional Imaging
Year :
2021
Submission ID :
1825742
Source :
www.aesnet.org
Presentation date :
12/6/2021 12:00:00 PM
Published date :
Nov 22, 2021, 06:50 AM
Authors :
Balu Krishnan, PhD - Neurological Institute, Cleveland Clinic; Frank DiFilippo, PhD – Staff, Nuclear Medicine, Cleveland Clinic Foundation; Guiyun Wu, MD – Staff, Nuclear Medicine, Cleveland Clinic Foundation; Imad Najm, MD – Staff, Epilepsy Center, Cleveland Clinic Foundation; Andreas Alexopoulos, MD, MPH – Staff, Epilepsy Center, Cleveland Clinic Foundation
Rationale: Subtraction ictal single-photon emission computed tomography (SPECT) co-registered to MRI (SISCOM) is a valuable non-invasive neuroimaging tool in the presurgical evaluation of patients with medically intractable focal epilepsy (MIFE). We demonstrated, in patients undergoing SEEG evaluation, that electrical stimulation of the seizure onset zone during interictal periods produces higher evoked responses in brain regions that show hyperperfusion during seizures [1]. Strong neurovascular correlations were observed between SEEG-recorded seizures and perfusion changes quantified using SISCOM in a cohort of 70 MIFE patients [2]. Our studies have systematically shown that SISCOM hyperperfusion patterns are not random and correspond to electrically connected brain regions that are activated during an epileptic seizure. Variability in injection administration, seizure duration and type influence the diagnostic value of SISCOM [3,4]. In this study, we systematically investigated and quantified the similarity between networks activated during repeat ictal SPECT injection.
Methods: Sixteen patients who had a repeat ictal SPECT investigation and were seizure-free following resective epilepsy surgery for at least 6 months were identified for the study. The patient's MRI was parcellated using the USC-Brain atlas and served as ROIs. The mean SISCOM z-score across all voxels within an ROI was calculated for both SPECT scans. Cross-correlation analysis was employed to quantify the similarity between ictal SPECT studies. Dice coefficient was used to quantify the similarity of hyperperfused regions across both studies. Subgroup analysis was performed to investigate the effects of ictal SPECT injection delay, duration of seizures, and seizure continuation time on the correlation between the two ictal SPECT scans. Mann-Whitney U-test was used to access statistical significance.
Results: Across patients, we observed a statistically significant correlation (r2=0.20, p< 0.001) between the SISCOM z-score of the two studies (Fig. 1A). These correlations were independent of the timing of ictal SPECT injection, duration of seizures, or seizure continuation time (Fig. 1B-D). Thirteen patients had a good quality postoperative MRI available for analysis. A higher number of patients had SISCOM hyperperfusion overlap with the area of surgical resection when the second ictal SPECT scan was used when compared to the first scan (85% vs 62%, p=0.19, Table 1). Finally, a statistically significant dice coefficient (DC=0.12, p< 0.001) was observed when comparing the hyperperfused ROIs across both studies.
Neuro Imaging