thank you hello my name is John Ralston and I am a functional neurosurgeon at Brigham and Women's Hospital and Harvard Medical School in Boston Massachusetts and this lecture is about the principles of stereotaxy or stereotactic Neurosurgery so one very common way of doing Neurosurgical procedures is to directly see directly visualize the target of the surgery and this is what people like Busey did in the 1930s when they were targeting the motor cortex for extra patient and Parkinson's disease or what Cobb was doing in the 50s when he was sectioning you fibers of the same area which is basically a multiple appeal transaction for Parkinson's others targeted different parts of the cortical spinal tracts like Walker targeting the cerebral peduncles and the paniculotomy in the 40s or Browder targeting the internal capsule with a capsule transaction in the 50s Cooper serendipitously discovered the role of the Globus pallidus or the extrapyramidal system the EPS when he accidentally like it at the anterior caroidal artery which was a great treatment for Parkinson's if you can really restrict that lesion to the Globus pallidus so the question really became how can you do this select targeting less invasively and this is the Crux of stereotactic surgery so we need to know where a Target is in an individual's brain from the outside and everyone's brain's a little different so you need to adapt to those idiosyncrasies and then we need a safe way to get there the etymology of stereotactic surgery is pretty interesting so Horsley and Clark initially called a stereotaxic surgery and that comes from two Greek Roots stereo which is Greek for solid or three-dimensional and taxes which is Greek for arrangement as in the word taxonomy so together these were stereotaxic which might mean something like 3D arrangement of structures stereotactic which is what we say nowadays is actually a mix of Greek and Latin and the Latin root actus means touch so here stereotactic might mean something like touching in three dimensions this issue came to a head in the 1970s that the precursor for the world Society of stereotactic and functional neurosurgery and they wanted to know what to use for that s in their name the journal confinia neurologica was the main journal for functional neurosurgery before then and that means boundaries of Neurology but they changed their name thereafter to stereotactic and functional Neurosurgery all these stereotactic procedures are a form of intracranial navigation and for this you need a map which shows you your Target and then some sort of fiducial or location that tells you you are here so some sort of marker that you can compare on the map to the real world the maps we use in functional neurosurgery are typically things like atlases the most famous probably being the Shelton branded Warren Atlas which used a few cadavers to take slices of the brain and then delineate important structures on them but also things like CT scans or MRIs which again have coordinates that you can look on those are either the dicom coordinates or coordinates that are related to landmarks you can see like the anterior posterior commissures the fiducials for all these are things like external landmarks so the nasion or the external auditory meatus the frames themselves which you can visualize in CT scans or MRI scans skin markers that you might put on the little stickers or bone fiducials which you can screw in before taking a CT scan or even surface registration where you use the patient's face nose and facial features as a way to register those to the images you're looking at all these stereotactic frames we've used uh in history have relied on this idea of a map and some sort of fiducial the first frame I could find was made by zernov in Russia in the 1880s and here this was used to drain a brain abscess using external landmarks the myelatoma trendellumber the same Trendelenburg who the positions are named after was actually similar to a copy machine that Thomas Jefferson had come up with in the 1700s and this used the actions you would perform in a fake brain in non-human primates to mimic into the real brain so here the the atlas was this fake brain and then the actions were copied into the the real world Horsley and Clark came up with a more complicated and sophisticated frame in the early 1900s and they used this to create lesions in non-human primates and here they're landmarks or their fiducials were external landmarks or craniometry other frame is quickly came into use like moosens in the 1910s although no one used it in humans Martin kirschner the same kirschner who did K wires came up with a human frame for things like percutaneous rhizotomies in the 1930s again using external landmarks but what really changed the field with Spiegel and vices frame in the 1940s and this was the most contemporary one at the time and it was used to ablate the thalamus for psychiatric disorders and here they used a combination of plaster of Paris casting and pins to keep the frame in place while it was on the patient's head lexell visited Spiegel and vices and shortly thereafter came up with his own frame in the 1940s the late 1940s and this was the first with The Arc quadrant system and by this we mean specifying a vector rather than just a Target so there's a Target that's defined in XYZ coordinates at the center of the the system but then there's a vector which is determined by these rings and Arc angles that gets you to that Target so the target can remain fixed but the approach angle the vector can change drastically depending on how you adjust those those two ring and Arc angles frameless systems came about later in the 1980s Roberts described a system that projected images from a CT onto a microscope and Watanabe used a An Arm based system that looked at the Joint angles of the arm to determine the position of tools in space and compare that to the the position within an MRI and this is the basis of things like the Rosa renishaw robot nowadays and of course now we're spoiled with all sorts of new techniques and Technologies so robotic neurosurgery intraoperative Imaging tools like the clear point system augmented reality virtual reality frameless stereotactic navigation and things like even focused ultrasound which is another stereotactic procedure but all these still rely on the same general principles going back to the idea of using a map we can Target things on these maps with two different ways the first is indirect targeting so here we cannot see or do not see the Target that we actually care about a good example of this might be the ventral intermediate nucleus of the thalamus which is difficult to visualize on most contemporary MRI sequences here you can rely on things you can see and then infer the location of your Target by measuring from those easy to visualize things classically we would use something like the acpc coordinates or the anterior and posterior commissure which are really easy to see on CTS and MRIs you can also use things like the nasion if you're looking for a coker's point or other external landmarks like that um direct targeting on the other hand so compared to indirect targeting direct targeting is where you can actually see the Target on the scan so something like the subthalamic nucleus on a T2 scan or the Globus pallidus on an F gate or or white matter null or mp2 rage why we tend to use acpc has an interesting history too and that's because when we used to use ventriculography that was a really easy structure to visualize on these skull films the anterior and the posterior commissure were very easy to look at and then you can see those compare them to your frame which is also in the same images and use that as a way of navigating so acpc space is often the way we reformat all of our images looking again for the anterior and posterior commissures the posterior commissure you can find by following the tectum to its tip essentially the typical distance of this is about 25 millimeters which is always a good check to look at that and record it and the mid commiserable point which tends to be the origin of most coordinate systems that use acpc space is that halfway point between those two structures also of note you need a third point for these to just delineate a three-dimensional space if you just have the AC and PC you have a line but that does not determine everything you need a third point above that in the center to really give you the the three-dimensional space reference system so when you have something like a CPC coordinates you can come up with indirect coordinates for a lot of structures we care about something like the the ventral intermediate nucleus of the thalamus the Globus pound this is the subthalamic nucleus but everyone's brain's a little different so there's ranges for these indirect coordinates these are good starting points but they're not always accurate for an individual to directly Target things we use images that can visualize these structures better so for the GPI for instance we can use an F Gator or a white matter and all or any sort of inversion recovery type sequence where you can see the border between the gpe and GPI and then Target the GPI directly or for the stn you can use a T2 or a swee scan or some sort of image that really shows you the stn really well and also the red nucleus is another good Landmark that helps you target within the stn accurately once you have these images and everything's targeted you need to get from that map which is in arbitrary MRI space or CT space and then get that to the patient-specific stereotactic space that you're using with your frame or robot or other system and you usually do this by registering your images so you have your pre-operative MRI for instance and then you have a CT scan with a frame in place and you need to register those because the CT shows you something that you don't see on the MRI which is the stereotactic frame or this ET might show you the bone fiducials you put in so you need to get these into the same space so you can do all your referencing based upon those fiducials in the real world you essentially need that you are here sign on your images so Fusion is a way of looking at two images and trying to match them as best you can there's a lot of automated ways to to do this perform this step and usually pretty accurate by just letting the software work but you always need to double check the software can often uh sort of misfire give you errors and it's up to you as the surgeon to validate these fusions and make sure that you trust them localization is another step up here which is how you localize the instruments you're using which is the frame in this case so localization usually uses some sort of visual cues to localize your your frame if you're using a stereotactic frame so this is an example of using the crw's localizer Box finding all these CT visible rods marking them on each slice and then using that to infer the location of that base ring for the crw so these localizer boxes typically use something like the N bar which was invented by someone named Brown who's a medical student at the University of Utah when he came up with it and here there are three points in every axial slice you obtain when this localizer box is on corresponding to where you might be on these ends as you slice through them and so you know the actual distance between the two outer bars of these ends and so when you see that on your slice you know how far that should be and then you also know roughly or can figure out where when you find that middle bar what slice you are on so if you see two of those parts of the the end really close together that might tell you that you're really high on the slice if you're really far down the other two bars might be close together because you're you're closer to the ring base and so using this in just simple geometry you can determine exactly what level you are so how high you are from the base ring and if you're at any angle because going at an angle will make those the the two most distant bars of the end appear farther apart because you're slicing at a different angle than true axial so bringing all these together you have a stereotactic space now which is defined by your frame or your robot or other instruments you're using and then you have a Target that you've defined and that might be in dicom space on an MRI you fuse those together so now you have a Target that's in your stereotactic space and that's what you're going to Target for your DBS electrode or biopsy needles that that Target that's in the real stereotactic space of your instruments that you can see in front of you as you're operating for the Lex L frame this is a trajectory so you have the XYZ Target and then the trajectory is that Vector of how you get there and that's defined by The Arc and the ring settings of the lexel or the crw frame and you can always adjust those at any time without affecting the target really just the approach angle that you adjust with any stereotactic surgery there are sources of error one of these is the Imaging particularly with MRIs there is a lot of inhomogeneity the magnetic field and distortions caused by the field strength which is worse with the higher fill strength magnets of the seven Tesla is worse than the three Tesla which is worse than the one and a half Tesla although a lot of these scanners have algorithms to correct for this somewhat um there are other sources of frame like or air because of the frame itself being just used and having some intrinsic errors or systematic biases you can try to overcome some of this by being precise in how you place the frame so always trying to keep it in line with the line between the anterior and posterior commissures and that means keeping it in line with the cantho miedo line which is also called the line of Frankfurt you can try to keep the frame over the target if possible this is particularly useful for things like focused ultrasound if you over tighten the frame you can start distorting it which can make it less accurate if you don't assemble it correctly obviously that can cause errors and the frame for the localizer box or the fiducial Box must be seated inverly properly on that frame's base ring and if it's not if it's lifted up even a little bit that can throw off all the rest of the localization so sometimes if you get hair caught in between the localizer box and the base ring that can be pretty devastating to your subsequent accuracy frameless methods are are great they are tend to be a little bit easier to do and you can use fiducials that are either bone fiducials like those you see in the bottom left of the screen here that are anchored into the bone and you can visualize them easily on CT scans you can use these skin sticker fiducials like shown on the top upper right which are visible in MRI really well or you can just use the skin itself so if you do surface tracing of the person's nose and orbits and forehead that can be matched and lined up to a CT scanner MRI which gives you additional way to get some sort of accuracy using these frameless systems is often camera based so there are two cameras which can visualize uh the little reflective spheres that you put on instruments or reference array both cameras will see a slightly different image based upon Parallax depending on how far they are from the Target and using this you can just use Simple calculations to determine where things are in the real world there will be some limitations to this accuracy the camera needs to see everything but this tends to be the Workhorse of a lot of modern stereotactic navigation we do for things like brain tumor resection you can also use electromagnetic radiation as a source of a new coordinate space so here if you have a cubital electromagnetic field a gradient you can use that gradient to determine the position of tools like you might be able to use the the visual location of those IR sensitive spheres when you place fiducials it's important to have some best practices in mind so you want to avoid linear configurations if you have them all in a straight line that doesn't give you as much information as if they're out of the same line and again that's because you know two points to find a line but not a plane so you need more points to Define planes and this real three-dimensional space so arrange them uh not in a line make their Center of their configuration close to where you care about so the closer they are to what you're targeting the better spread them out use as many as you as you can reasonably Mark the scalp when you do the fiducial if you're using the stick on ones because they can fall off and often do and then you want to avoid the occipital region or any part of the scalp that might be moved so when you position a person that occipital skin in the back and fold and then change where the location of the fiducial is if you put tegaderms or other stickers over the eyes that can distort the the eyes if you have a tracheal tube in that can distort the face as well or the nose so be cognizant of these things if you're using those surfaces to register when all is said and done the accuracy of these various methods of registration has been studied very much in the past bone fiduciials have great accuracy around a millimeter skin of fixed fiducials are less accurate as you can imagine surface matching is probably about five millimeters of accuracy sometimes worse sometimes better a frame is usually one to two millimeters maybe three millimeters things like the the star fix are about one to two about the same as the frame same with the next frame Clear Point the interoperative MRI one is probably one of the most accurate less than a millimeter because it is being constantly imaged and compared to the MRI or the the target directly within MRI space so again more sources of air which are things we we dread in stereotactic surgery again the Imaging itself is a source of air a lot of these scans have a resolution of uh you know millimeter voxel size so you're not going to get better than that if that's your image resolution there are inhomogeneities in the field strength for MRIs like we talked about there can be artifacts created by the frame or other things that are nearby image Fusion creates some air so if you are using the the frame as a fiducial in the real world and your image Fusion error is a millimeter then your Target's going to be a millimeter off too there's registration error so if you're using a surface registration for instance that can have a new source of error for you and then there's effector error so the the frame itself might be distorted or bent the frameless system might have some intrinsic construction defect that causes it to be a little off there's going to be some manufacturer tolerances which are going to cause a little bit of air and lastly there's you as the surgeon which is another source of air so if you mismeasure something that's going to cause problems and this is particularly acute in the frames where you have a lot of measurements to make so the lexl frame there are about eight measurements you do there's typically two x's two y's two Z's one Arc and one ring measurement totaling eight measurements you have to make for every trajectory and let's say that you have a one percent error rate that means you're going to have an error every you know six or so DBS cases when you're doing bilateral trajectories or if you're doing you know 12 leads in an seg case and using something like the crw frame that means you're going to make an error every case so it's important to check these measurements and not just yourself but whoever's assisting you because humans make mistakes so to conclude this short lecture stereotactic surgery uses some externally visible Landmark also known as a fiducial and that could be something like a frame or a bone fiducial something you can see in the real world in the operating room to find an internal Target based on some map and that's either an MRI or an atlas or some other way to find what you're seeking within the brain deep in the brain with the goal of being precise and minimally disruptive to non-targeted parts of the brain so that is the the Crux of stereotactic surgery and that's how we're doing it nowadays and there are many resources that are excellent to give far more detail than we've discussed here like these two books that just came out recently so not a paradian Samir chef stereotactic and functional neurosurgery or Bob gross and Nick bullis's Neurosurgical operative Atlas both excellent resources among Myriad others out there so thank you for your attention Amanda if you have any questions please don't hesitate to contact me [Music]