Next, we have Dr. Vishwanathan speaking about asymmetric development of lesions from stereotactic radiosurgical capsulotomy for refractory OCD. Thank you very much, and good afternoon. My name is Vignesh, and I'm a student who has been working in Dr. Sameer Sheth's Functional and Cognitive Neurophysiology Lab at Baylor College of Medicine and Columbia University for the last couple of years. Today, I'll be talking about stereotactic radiosurgical capsulotomy in the context of obsessive-compulsive disorder, in addition to some findings about the way lesions develop as a result of this procedure. I have no disclosures. OCD affects 2 to 3% of the world's population, and most cases can often be treated pharmacologically or behaviorally. However, for a number of patients that have severe and refractory cases of OCD, neurosurgical intervention may be an option. One such procedure is stereotactic radiosurgical capsulotomy, or SRS capsulotomy. And SRS capsulotomy is a bilateral procedure with identical radiation prescription dosing using gamma knife, and the maximum dosage usually ranges from 150 to 180 gray. And lesions are created in the anterior limb of the internal capsule. Studies have shown this treatment option to be a viable method for reducing OCD symptoms, yet there have not been any studies looking into post-operative capsulotomy lesion development. And this is what we set out to answer. We wanted to see how these SRS capsulotomy lesions develop bilaterally in OCD patients. More specifically, how do these lesions vary in size, and how do they vary in space? And are there ultimately any systematic or consistent ways in which left and right lesions differ from each other as a result of the procedure? To go about conducting the study, we used scans from 38 patients from University of Virginia, Brown, and Sao Paulo, for which there was one scan per patient. The lesions were masked by two radars to control for the credibility and accuracy of the lesion volumes. And for subsequent analyses, we used the average of the two radars as raw lesion volumes for relevant statistics and calculations. The volumes were then measured in the patient's native space. And for the spatial analysis, we non-linearly transformed all of the lesions to one millimeter MNI space to ensure standardization of all patients. The overall method that was implemented to examine lesion asymmetry from both the volume and spatial standpoint was descriptively quantifying the differences, creating an asymmetry index to keep track of those differences, and then testing for systematic or consistent left to right sided differences. For the volumetric analysis, we were interested in differences of lesion volumes across individuals. So the left and right sided lesion volumes were compared, a mean asymmetry index was then created, and then a paired t-test was run to determine evidence of systematic or consistent asymmetry. A parallel approach was implemented to compare lesion location between the left and right sides across patients, and we labeled this the spatial analysis. Left to right sided spatial differences were quantified by reflecting the right sided lesions onto the left side, and a heat map was subsequently generated. The Sorensen Dice Index, or the Dice coefficient, was used as a means to create an asymmetry index to evaluate the degree of spatial similarity. And then a general linear model was then used to run a statistical test evaluating for systematic differences. So these are the initial results of the volumetric analysis. And the left side shows the means of the left and right lesion volumes, and their respective standard errors. So on average, the left and right sided lesion volumes seem to be fairly similar. And on the right, we show connected points representing individuals as lesion volumes. However, this shows that the left and right side are fairly similar, sorry, fairly different. And given the decent slopes that we see indicating that individually, left and right sided lesion volumes differ. We then created an asymmetry index, which for the volumetric analysis was the ratio between the volumes of the smaller and larger lesions. The points represent the ratio of smaller to larger lesions for each patient. And we see that the mean asymmetry index is .55. And this means that for the average patient, the smaller lesion is about half the size of the larger lesion, suggesting substantial asymmetry at the individual level. Now if there was no asymmetry, then we would expect to see a more uniform distribution of the data points. But instead, we see this bulge occurring around .55. In order to determine if there was a systematic or consistent volumetric preference for the left or right side, a paired t-test comparing left and right lesion volumes was run across the entire cohort, and the difference was insignificant. So up to this point, we have seen that capsulotomy lesions in OCD patients are laterally different on an individual level, given the average asymmetry index. However, when comparing the volumes bilaterally across the entire cohort, the paired t-test showed that systematically, there is no significant difference. And the spatial analysis was approached in a similar fashion. Lesions were nonlinearly transformed, and the spatial averages of the lesions were computed. A heat map was generated with the red representing the left side and the blue representing the right. And this shows that similar to the volumetric analysis, the spatial averages are fairly similar. To quantify differences, we reflected the right lesions over the left, and we used the Sorenson-Dyson index as a metric to quantify spatial differences between the left and right side of the lesions. And the SDI is the volume of the overlap over the volume of the average, and on the right, we see that the SDI plot where the range of the index is from 0 to 1, where 0 represents no overlap and 1 represents perfect overlap. We calculated the mean SDI as .32, so on average, there is about a third, one third of spatial overlap. And on an individual level, this indicates that there is substantial left-right asymmetry from a spatial perspective. Once again, if there weren't spatial asymmetry, then the bulge in the figure would not be there. We would instead, once again, see a more uniform distribution of the data points. And having found substantial asymmetry within patients, we wanted to see through a GLM if there was a consistent location preference between the left and right-sided lesions. So a voxel-by-voxel t-test indicated that there was no significant difference, suggesting that the location of the left and right-sided lesions is not preferred in a consistent way. To summarize, volumetric and spatial differences of right and left-sided lesions of SRS capsulotomy are quite different within individuals who have refractory OCD, yet across the entire patient cohort, there seems to be no manifestation of a systematic left- or right-sided preference. And overall, the findings suggest that the volumetric and spatial development of these capsulotomy lesions at high radiation dosages varies in an idiosyncratic manner within individuals that is likely multifactorial. And these dosages are orders of magnitude larger than those typically used for most gamma knife treatments, so subtle differences in the radiobiological response will be amplified. For future work, we plan on looking into examining capsulotomy treatment variables, and more specifically, order of shot delivery. Patient variables is also another area that we would like to look into, such as gray-white matter fractions. And looking at these lesions developed through time is another area we would like to examine to see if this could help us understand more about this asymmetric evolution. Many thanks to Dr. Sheth for the opportunity to work in his lab and for allowing me to take on this project, in addition to his constant support and encouragement. Big thanks to Pranav Nanda, Dr. Pathak, and Dr. Banks, and all the other esteemed co-authors at UVA Brown and Zappalo. And thank you to AANS for the opportunity to present as well. Thank you very much. Any questions? Can you try the microphone, please? Just that close. Take that for a note. Is it working now? No. I'm not doing sound, I don't know. So, at least you can hear me. Yeah. Let's repeat the question of Dr. Regis to the audience. For sure. Congratulations. Thank you. It was a very nice presentation. These cases are coming from several centers. Yes. And if you, as we look at the distance between the two shots can vary depending on the patient Right. neurosurgeon doing the planning. Have you tried to correlate the variation in size of the image of the MRI with the variation in size of volume of the planning, the 50%? Right. I miss it in your talk. No, we haven't looked into that yet, but it's something that we'll definitely take into account. Maybe the first thing to look at. Yeah, for sure. Yeah. My question is similar to Randy. These are different, they take different centers. And they use different number of ISO centers on both side. And some, for example, were staged at different times. For example, the Brown work, many of the cases were done with wide staging between one side and the other side, while others were done simultaneously. So how did you control for the variation and how the technique and the size of the ISO centers were done? So the standardization, like I mentioned before, for the lesion location, we standardized all the scans. We non-linearly transformed them into one millimeter MNI space. And that was one way we controlled for the lesion location. But yeah, that's something we'll definitely have to take into account for future analyses. And did it matter to the white box score? No, I'm not really sure, actually. Thank you.

stereotactic radiosurgical capsulotomy

refractory obsessive-compulsive disorder

lesion development

individual-level asymmetry

evolution of lesions