Temporal lobe

image with frontal operculum removed

8 - temporal stem, limen insulae 

Surface anatomy:

6: squamous suture, boundary between temporal and parietal bone. Half of the temporal lobe is actually above it... 

1: coronal suture, goes right above anterior temporal tip. 

2: precentral gyrus

3: postcentral gyrus 

when you split sylvian, if you retract too aggressively on anterior frontal lobe you can cause weakness. 

12: pars orbitalis

8: pars triangularis

9: pars opercularis 

4: precentral

5: postcentral 

2: central sulcus 

6: supramarginal gyrus 

superior & inferior surfaces of temporal lobe:

White matter: 

Occipital frontal fasiculus = IFOF 

ST= striae terminalis

T= tapetum 

Temporal stem:

Uncinate fasiculus - front of temporal stem 

Hangman Fractures

bilateral C2 pars fractures that can result in C2-3 spondy; classically due to hyperextension and axial loading, but there are multiple possible mechanisms 

These are the levine-edwards classifications, and they are stratified by mechanism of injury 

(A) Type I: hyperextension, axial loading - hairline fractures. Stable, non-op, collar for 4-6 weeks. 

(C) Type II: hyperextension, axial loading + bounceback flexion that can tear PLL/C2-3 disk. depending on how bad the spondy is, you can treat in collar +/- closed reduction, or fuse.

(C) Type IIa: hyperflexion + distraction: the key is you want to apply compression and some extension - i.e. compression halo or neutral collar 4-6 weeks +/- compression closed reduction. Do not distract these people. 

(D) Type III (rare): type II injury with bilateral dislocated facets. Needs reduction + fusion. CTA to look at verts. You may need to operate through aspirin. 

Jefferson Fractures 

Compression-type fractures that can be symmetric (burst from pure axial loading) or asymmetric (fractures biased to one side from hyper flexion/extension/lateral flexion) 

These are typically nonop as long as the transverse ligament is intact. Some people don't even collar these people if the fragments are not that displaced. If the transverse ligament is compromised though, you must operate, most likely O-C2. If the fracture morphology is amenable, C1-2 fusion only may be possible. 

MRI of disrupted transverse ligament 

Image sources

http://www.radiologyassistant.nl/en/p49021535146c5/spine-cervical-injury.html

http://pubs.rsna.org/doi/abs/10.1148/rg.2015150035

C1 and C2 screws

C1 screws

C1 from the side looks like a pan with a handle, the pan is the lateral mass and the handle is the ring.

Dissect down til you find the ring of  C1, then drop down along the ring deep to the lateral mass. The C2 nerve root is encased in a bunch of fascial tissue and surrounded by a venous plexus that bleeds a lot, very quickly. You can lose hundreds of CCs of blood before you know it. This is not the vert, its venous bleeding -- the vert is typically going to be lateral to you at this point - although always look at CT/MR beforehand and make sure no anomalous vert paths. 

The vert runs lateral to the lateral mass, then goes around and up along the superior aspect of the ring of C1. People say that as long as you are at the midpoint of C1 or more inferior, that you are safe- this is usually true but not always. People say you are safe as long as you are within 1cm of midline. this is usually true but not always. Always look at imaging ahead of time and measure and figure out where the vert is. 

Retract the C2 nerve root down. Your entry hole should be midpoint (in medial-lateral direction) along the lateral mass, and as superior as you can get (in superior-inferior direction) - see the red dot in CT above. Aim approx 10 degrees medial. This is because you really don't want to go lateral, as you risk hitting vert. Use fluoro to localize your angle superior-inferior, aiming for the anterior ring of C1. The temptation is to go high, because the ring of C1 may overhang down and push your hand down. if you go too high you might end up putting the screw through the O-C1 joint. You may have to drill off the inferior part of the ring of C1 to get the optimal angle. 

C2 screw 

There are three kinds of screws you can do: pars, pedicle, and lamina. 

Pedicle screws are the strongest, but also the highest risk of causing injury to the vertebral artery. They do these a lot in Japan. Possessing balls of steel is a pre-requisite for doing these. Pars screws are very strong and very safe as long as you stay behind the PLL.  Laminar screws are very strong and low risk but a pain to connect to the rest of your lateral mass screws with rods because of the angle; some people do these primarily, but most do them as a bailout if your pars screws break out. 

see the first CT in this page, see where the vert is. As long as you are behind PLL, you are almost always safe. 

When placing C2 pars screws, dissect down to where you can see the medial border of the ring of C2.  your entry point should be 2mm lateral, you aim slightly medial to avoid vert and superior. If you put an instrument between the facet of 2 and 3, it will give you the exact angle you need to aim in superior-inferior direction. 

Helpful videos:

https://www.youtube.com/watch?v=bqu8okFUUcU

Sublaminar wires

The principle is you put some sort of bony graft (i.e. iliac) between C1-2 and use cables to hold C1-2 onto the graft until bony fusion is achieved. Depending on the technique used, the rate of psuedoarthrosis is high. Needless to say, only works in instances where both C1 and C2 are structurally intact, and where decompression/lami is not required. 

Sonntag technique - iliac crest jammed between spinous process of C2 and the interlaminar space of C1-2, wires go under C1 lamina, and loop under C2 spinous process. 

Retrosig cranis!

Indication: lateral posterior fossa structures; with a big retrosig you can reach all the way from foramen magnum below (ie to open cisterna magna for CSF drainage) to the back of the sella above, and to lateral clivus anterior (maybe even further, depends on how big the pre-pontine cistern is and how tight the posterior fossa) but the optimal working window is gonna be around the CPA. Workhorse approach for CPA pathology, MVDs to access CN 5-10, and tumors that are either very lateral and/or very deep - i.e. where you would rather retract cerebellum medially than punch through it.

Before the case, look at imaging:
- location of transverse and sigmoid sinus relative to EAM/mastoid
- thickness of mastoid temporal bone
- amount of air cells and how far posterior they extend
- slope of petrous bone (to make tiny adjustments to head position)

Positioning: dealer's choice
- supine + head turn: advantages are that it's a lot easier to set up, so if you are trying to do a lot of cases in one room and don't want to lose an hour per case on positioning, it's a good approach. The other mostly hypothetical advantage is that gravity pulls the cerebellum down and out of your field, but in reality most people have a pretty full posterior fossa, even with generous CSF drainage you end up having to do a lot of retraction and the gravity assist thing doesn't make a big difference. The main disadvantage is that for all the intradural work, you have to operate UP - which means sitting down with your arms up high in front of you. This is a hard position to hold, and most people will develop muscle fatigue in their hands and forearms pretty quickly, even with a chair where you can rest your arms. This works great for a procedure you can do in 15-20 minutes (i.e. MVDs if you're good at them) but this is not gonna work for a big tumor where you may need to cusa or drill for hours and hours. And since the microscope is often positioned very angled or lateral, the observer scope can be in a difficult position, which means the first assist often has to assume an incredibly awkward half-stoop. NB: Asking your ENT colleagues to drill the petrous in this position is a fast way to make enemies.
- 3/4 prone - tedious to set up, but much easier ergonomically for long cases. The goal/final position is head facing halfway between lateral and prone - i.e. 3/4 prone. You stand behind the patient's head, so that when you are moving in the corridor between petrous and cerebellum, you are going essentially straight in, which is nice and ergonomic for microscope neurosurgery.

How to position someone in 3/4 prone - there are many variations
- reversed OR bed, put the head of the bed on.
- put beanbag on OR bed before you have patient get on it.
- anesthesia intubates and gets all their lines. tape the eyes shut. if you are using neuromonitoring, they should get their needles in now and tape them in well
- put the adaptor on the bed that secures the upper-arm holder device
- pin - the goal is to have the bar of the mayfield roughly parallel to the floor in final position. Pinning you put two pins low on the occiput,  and contralateral goes high on forehead behind hairline if you can. The further your incision from pinna, the further towards contralateral ear you will have to move your posterior pins and the closer towards ipsilateral ear you move your front pin.
- first, position patient in lateral. The idea is that both shoulders/arms will be off the bed in the end, so you need to slide the patient up in the bed until both their axillae are above the head of the bed, which you will remove and replace with the mayfield holder. Then you slide the patient over to one side and push into lateral.
- put an axillary roll under their chest at the edge of bed to create space in the axilla
- suck air out of beanbag to hold them in lateral.
- now all at once (and you need multiple people to help you on this step) - you remove the head of the bed to exchange for the mayfield holder, at which point both the patient's arms/shoulders will no longer be supported by the bed - and thus require people to hold them. The dependent arm, cradled, goes into a soft foam sling that secures around the contralateral shoulder. Put foam padding under the straps. The superior arm goes into some sort of arm support device (we use a metal frame with a stockinette over it) that secures to the bed.
- turn the head to 3/4 prone, and lock the mayfield.
- the superior shoulder should be tape-pulled inferiorly; when someone pulls with the tape, another person should provide counter-traction proximally so you don't york on the brachial plexus.
- if you are doing BAERs, no xeroform in ears.
- if you are doing facial nerve monitoring, now is the time to put it in -- blue into orbicularis oculi (stick into eyebrows), red into orbiculars oris (stick into skin next to lips), and the other two as ground into the shoulder. NB: when you put in needes, put them in at a very shallow angle-- if you stick straight in you risk the mucosa.
- if you are using navigation (i.e. to find the sinuses) now is the time to register

Incision
- for a tiny quarter-sized retrosig (i.e for MVD) two fingerbreaths behind pinna, approx from digastric notch inferiorly to top of pinna superiorly
- for a big retrosig for tumor, four fingerbreaths behind pinna, from skull base to four fingerbreaths above

Prep, drape the ear out of the field.

Knife through skin, bovie to bone with impunity. you can do this part fast, there's nothing at risk
When you dissect periosteum off the bone, you can do so with impunity posterior, but watch out anterior - you dont want to dissect through skin into EAM. Poor form. So when you move anterior just go slow and feel with your instrument (pen 1 or joker or whatever) - you will feel the bone drop off into the EAM before you hit it. If you feel that, stop. its Ok to find that ridge as long as you don't violate the overlying skin.

When you retract the skin, if you are pulling skin anterior - warn your neuromonitoring techs, because you may disloge the ear needles they need for BAERs.

Now you have to figure out where to drill. Where are the sinuses? At this point, the head is in 3/4 prone and buried underneath yards of drape. you cannot rely on typical craniometric landmarks - you won't be able to feel inion or zygoma or tragus.

The sigmoid sinus will run essentially right behind the EAM. Remember, deep to the sigmoid sinus you will find petrous bone, just as deep to transverse sinus you will find tentorium. The transverse sinus you can find by the curve of the skull - as the skull curves from the flatness of temporal bone towards the curve of the sub-occiput, that junction lies the transverse sinus. There also lies the attachments of the cervical musculature, and thus the junction between cervical fascia and galea.

Or you can use navigation.

blue - transverse and then sigmoid sinuses

pink - where you drill for small crani

dotted pink - how you expand your drilling for a bigger crani 

black - pinna and EAM

green- incision for small crani 

Drilling
- beforehand, you looked at the CT scan to see how thick the bone is, and also to see how far posterior the mastoid air cells are. When you are drilling, if you run into something that looks like dura really early on, either they have a super thin mastoid temporal bone and you're running into dura, or its a mastoid air cell.
- you want to eggshell out over dura and sinus -- sometimes, the sinus can push upwards and will actually be more superficial than the dura. So if you've found the dura and there's bone over the sinus, you cannot just drill with impunity to the same depth over sinus. You might run into sinus before you expect. When you've eggshelled over dura, peel off remainder of bone with curette and kerrison (slowly and carefully anterior and superior, quickly and with reckless abandon posterior and inferior). When you get close to sinus, use a woodson, point it upwards flush along the bone, and sweep side to side carefully. You will feel the sinus. Don't jab or poke it with the instrument.

Wax the mastoid! Wax in and wax out, as one of my wiser chief residents once said. You cannot wax too much! if your patient leaks postop you will regret not having waxed more!

What to do if you hit the sinus
- step 1: do not panic.
- step 2: tell anesthesia that you hit the sinus. A decent anesthesiologist will understand that this means (potential) massive blood loss and air embolus risk.
- step 3: if lots of blood is coming out, stick your finger over the hole to stop the bleeding
- step 4: obtain a very large (very large!) gelfoam, and put the gelfoam over the sinus injury, put a patty over the gelfoam. Do not use a small gelfoam, it could get sucked into the sinus and cause a sinus thrombosis or a PE
- step 5: wait, and reassess. sometimes, waiting it out with gelfoam + pressure is enough for it to stop. If not, you may need to think about a dural flap closure, so you may need to do more drilling - sometimes people even extend the crani supertentorial to obtain a patch of dura to flap down over the sinus.
- step 6: call for help early from a more experienced surgeon

Open the dura in a c-shape with the flap side towards the sinus. Some people like knife through outer leaflets of dura, then woodson + knife, some people like tenotomy scissors with geralds to pick up dural edge. Use a patty to protect the brain no matter what you do.

Tack up the dura with neurolon sutures. Put pattys or telfas on the brain, and with gentle pressure from your retractor, suction out CSF slowly to relax the brain. if you have a lumbar drain, pull off 15-20 cc of CSF from it. In a big crani for tumor, direct these efforts southward - you don't need to do a crani all the way to foramen magnum, just far enough to reach with gentle pressure to open the cistern.

In an MVD for TN, direct your efforts towards the junction of petrous bone and tentorium ("the tunnel of love"). FYI you will often run into some sort of superior petrosal vein superficially. sometimes there is one, sometimes many branches. Try really, really hard not to sacrifice it - the complication rate described is small but not insignificant.

How to place a subdural drain

NB:
- this only works for chronic or mostly chronic subdurals (a few streaks of acute or some layering subacute is OK)
- if you think there is a good chance you'll have to convert to a crani, just do it in the OR instead

Entry:
- For 80-90% of subdurals, the optimal entry point will be frontal, slightly above temporalis (you don't want to be cutting through that at bedside) - to find it, ask patient to bite down, feel the temporalis, and feel head for superior nuchal line. I like to go right above that.
-Enter right around coronal suture. I like going 1cm anterior to coronal suture - you can go on it or behind it. Look on scan beforehand for membranes; where the dura attaches to the suture often has a membrane, which can represent thicker dura and higher chance of bleeding.

Shave from entry point to 4-5 cm behind
Prep, lido all around tract, prep again, drape

Incision. Retractors.

Drill--
- some people drill perpendicular, and some drill biased aiming in goal direction.

Open dura with preferred method.

Passing catheter (we typically use EVD catheters)
- some people just put in the stylet in the direction of their goal, +/- biased drilling, and soft pass the catheter over it.
- I prefer to put ~70-80 degree bend on tip of EVD catheter, then insert through burr hole in an angle manner. Then, you soft pass over the angle. This way, you can control the directionality of the catheter.

Insert catheter ~5cm in. The important thing is to stop if you feel resistance, even if you are only 1-2 cm in. If you don't feel any resistance and you are confident in the direction of the catheter, you can go 7-8 cm in. Remove stylet while holding catheter tight.

As much as possible during the procedure, keep a wet raytec over the burr hole to prevent sucking in air. This is especially important when you are aspirating on the catheter within the burr hole.

Attach syringe (raytec over hole) aspirate ~10cc to make sure you are in. Tunnel EVD, suture closed incision, then aspirate however much you want -- in old people, probably don't want to take more than 40cc, old people brains don't tolerate massive shifts. Young people, ok to be more aggressive, can even put suction bulb on subdural drain.

Turning a pterional crani

- For aneurysms: Mannitol, decadron at beginning of case. Neuromonitoring with SSEP/EEG. For these cases, anesthesia is ok to paralyze. You should be in burst suppression by the time the dura is opened. When anesthesia hears your drill, they should start pushing the propofol.
- For tumors: variable. If operating near motor strip, we will often do MEP (if tumor is near corticospinal tract) and/or phase reversal (if you need to localize motor strip for cortisectomy planning or if tumor is near surface), no paralytics.
- If using stealth, cannot do any movement that break the bed after registration (i.e. cannot do head of bed up, can only do Tberg/reverse Tberg/airplane). When you raise or lower the head of the bed or the feet, it can make your stealth inaccurate by up to a cm. Doesn't make sense because the head is in pins locked into the mayfield, and the stealth arm is also locked into the mayfield, but this is the way it has been shown to be. (shrug).

- Supine, operative side facing out, ET tube on anesthesia side, tape eyes with tegederm (wait until after registration if using stealth - typically necessary for tumors and distal aneurysms) xeroform in ears.
- Pin with two ipsilateral, one contralateral -- something like picture below, but I typically like to pin a little closer to ear -- (see two blue dots) one superior and posterior to middle of pinna, one a few inches posterior and superior to that. Contralateral pin lateral above eye, close to hairline (green dot).

- Position head so malar eminence (pink dot) is highest. this will require rotation, extension, and significant translation of head superior relative to neck. 

- Incision from 1cm anterior to ear, straight up vertically until you are at superior temporal line, then slope over behind hairline to widow's peak. stay behind hairline. The sooner you slope over towards the face (i.e. if you were to go directly behind hairline) the more frontal lobe exposure you will get, the more you stay straight (or if you were to even swing back over the ear towards the back of head) the more temporal exposure you will get.

- Over non-temporalis: knife to bone.
- Over temporalis: knife through galea, bovie through temporalis to bone. Bovie temporalis below superior nuchal line so you leave a cuff of fascia on the bone flap.
- Position periosteal at root of zygoma, push up to get temporalis off in one flap. Advance flap until orbital rim.
burr holes, in order

Drilling
- The more burr holes, the easier it is to turn the crani. But the more (potential) cosmetic defect. In younger people with good dura, you can get away with fewer, but in old people with crappy dura that's all stuck to bone do at least 3 and do good dissection with woodson and penfield 3.
- You're not a badass if you can turn the crani through one burr hole if you're driving the strugglebus for 20 minutes with the footplate or if the dura looks like shit afterwards.
- 1. root of zygoma (blue dot)
- 2. modified keyhole (see photo) -- this is going to be slightly more superior than traditional keyhole. put drill on the spot I put my finger, and aim towards frontal lobe with perforator, you will end up on the red circle, directly on top of orbital rim. The goal is to make your crani with your footplate flush with the floor of the anterior fossa. That way you can get under the frontal lobe. If turn your crani too high on the forehead, you will have to rongeur down the frontal bone until you get the floor of the anterior fossa, and you will make an irreparable bony defect right on the patient's forehead, which will look bad, and they will be really pissed and think you are a shitty surgeon, which you are if you do this. To find the floor of the anterior fossa: feel the thickness of the orbital rim on patient, that distance represents the thickness of the rim all the way over the eye. Right above that will be the floor of anterior fossa.

3. Some spot inferior to superior nuchal line (so that it will be buried under temporalis) superior and posterior to burr hole #1.


Grey represents the rough locations of the frontal and temporal lobes -- the pterion is right on sylvian fissure. The more anterior you turn your crani, the more frontal lobe exposure you get, the more posterior and inferior you turn it, the more temporal lobe exposure you will get. Do you see how if you want more temporal lobe exposure, you angle your incision more posterior? So if you're doing a middle fossa crani your incision is posterior auricular, straight line up and down.

Sorry for my shitty handwriting

The top image is the order to drill - from superior temporal burr hole to keyhole (hypothetically less risk of getting into frontal sinus if you come from back to front), then inferior temporal burr hole to superior temporal burr hole. The further you curve this posterior, the more temporal lobe exposure you will get. Then come from the keyhole backwards towards your inferior temporal burr hole until you hit the sphenoid wing, then go the other way. you will likely need to curve superiorly as you drill near the sphenoid wing, thus getting a little heart shaped crani. You can crack the bone flap to get it off, but be careful - sphenoid wing connects to anterior clinoid, and you can propagate the fracture to the skull base, resulting in damage to optic nerve or carotid!

Then you need to drill off the remaining sphenoid wing until you hit the meningo-orbital band. Most people will retract the dura gently and/or put an instrument between drill and dura to protect it (i.e. penfield 1 or 3) however be careful, you can 100% cause contusions of the brain through the dura if you push on it too hard, and if you're on the left side this is prime real estate (brocas/inferior central sulcus/STG).

As an additional slick move, if you bevel the frontal/temporal bone (i.e. smooth out inner corner) you can get extra visualization/more retraction of dural flap.

Cut the dura in a c-shape with the flap/hinge at the sphenoid

Splitting the fissure
- knife through thickest arachnoid
- use fukushima suction to retract in your left hand
- use microtip bipolar in your right hand, insert tips closed, then open them to spread the arachnoid tissues

Anterior clinoidectomy

To access proximal ICA aneurysms (i.e. ophthalmic) you will typically need to drill off the anterior clinoid. Note: this is very high risk for injury to optic and ICA. The morbidity of these procedures is the reason why pipelines were invented.

1. Drill clinoid over the optic nerve until you unroof it. Use a diamond drill (cutting drills are too high risk of injuring surrounding structures) - your assistant must irrigate very aggressively, as the heat from the diamond is enough to cause thermal injury to nerve. This is very risky and requires very good drill control. 

2. Drill across top of clinoid from SOF (which is lateral; SOF is actually in the middle fossa). Previously, you should have dissected arachnoid/dural attachments off until you are able to reach the SOF.

3. The optic strut is a vertical piece of bone that lies between the optic canal and the SOF. If you drill straight down (where the circle'd X is) you will be drilling through the optic strut. you typically won't be able to drill all of it, and will have to crack it off. 

There is usually a loop of dura over the ophthalmic - this is the falciform ligament, and usually needs to be cut. Sometimes you have to cut the distal dural ring as well. 

Helpful images for paraclinoid anatomy

Spinal cord stimulator 

Make sure you know the history of all previous spine surgeries-- especially previous lamis. Look at the imaging. 

MAC anesthesia

Prone with back as flat as possible where you plan to enter. 

- C-arm in AP view, go live and move it around until you see the interlaminar spaces clearly 

- Go in two levels below where you plan to enter -- Put the coude needle on the skin and use xray to localize.  

i.e. if you want to enter T12-L1 interlaminar space, put needle in near pedicle of L2 (red), dock on lamina of L1 (blue), and enter the T12-L1 interlaminar spac (green) 

- needle goes in bevel UP

Ways to determine you are in the epidural space

- using your hands to get the haptic feedback of popping through ligamentum flavum, stop after you feel that. Downside: requires skill and experience. 

- syringe full of air that you periodically push against until you reach loss of resistance. Downside: air does not conduct, if you inject too much into the epidural space you won't be able to trial the electrodes during the case. 

- syringe full of water that you periodically push against until water starts to enter. Downside: if clear fluid comes out of needle, you won't know if its water that you injected epidurally or CSF

- Aim for direct midline placement

- Plug the electrode into the white box then extension leads. Wake up the patient intraop. The reps trial different settings and figure out if the pain area is covered. 

- Pull the coude needle out, tape the leads onto the skin and 

Pediatric neuroblastoma

- 3rd most common pediatric tumor after leukemia and primary brain

- originates from "neuroblasts" - essentially precursor neural crest cells, which make multipotential progenitor cells that give rise to the peripheral nervous system including dorsal root ganglia, the enteric nervous system, sympathetic chain, pigment cells, Schwann cells, adrenal medullary cells, cells of the craniofacial skeleton (including odontoblasts) -- within the brain, neural crest cells form nerve sheaths of cranial nerves 5/7/9/10, the optic vesicle (i.e. may eye structures like iris/cornea, as well as skeletal muscle that attach the eye to the skull), otic capsule/inner ear
- you can derive a neuroblastoma anywhere you have any of the above cells; most common is adrenal medulla in the abdomen, and along paraspinal sympathetic ganglia.
- depending on the degree of differentiation of the tumor stem cell, the resultant tumor will be  neuroblastoma, ganglioneuroblastoma, or ganglioneuroma in order of least to most differentiated. as you would expect, prognosis is better the more differentiated the origin cell line.
- paragangliomas i.e are tumors derived from chromaffin cells, which are differentiated neural crest

- 50-70% are metastatic by the time they are found - liver, lymph nodes, bone. This is because neural crest cells are undergo an epithelial to mesenchymal transition early on in their development,  and thus acquire enhanced migratory abilities and decreased requirements for intercellular contact. this allows the cells to leave the neural tube and go about the body.

- CNS mets are not very common, but when they happen, its nearly always to bone - calvaria and skull base. Especially at junction of lateral orbital wall and greater wing of sphenoid. They can also metastasize to dura and will typically expand along exterior aspect of dura between dura and bone. They have a characteristic "sunburst" appearance --- ddx also includes ewing's

- as treatment for neuroblastoma improves, the rate of CNS involvement is increasing - this is likely because kids are living longer with the disease, and the chemo that they get that achieves tumor control may not have good BBB penetration, which means the CNS may be a sanctuary site for the tumor. 

- very rarely will metastasize to brain parenchyma, into ventricles, or leptomeningeally - parenchymal mets often will bleed or be calcified, qualities which are rare among pediatric brain tumors 

- Imaging - they show a sunburst/hairlike projections of bone into tumor -- this is from rapid, aggressive growth of tumor causing a periosteal reaction. Other tumors that do the hair thing - ewing's sarcoma (used to be believed that they had a neuroblast origin cell - actually its a fusion gene expressed in mesenchymal stem cells), other sarcoma, lymphoma/solid leukemia

Helpful papers: 

https://www.ncbi.nlm.nih.gov/pubmed/19353343 - Calvarial lesions: a radiological approach to diagnosis.

https://www.ncbi.nlm.nih.gov/pubmed/20410407 - Imaging of metastatic CNS neuroblastoma.

Visual system

Optic Tract 

LGN

- The vast majority (~80%) of retinal ganglion neurons terminate in the LGN, which is a dorsal thalamic structure.
- BUT 80% of the excitatory input to the LGN comes not from retina but V1 and other cortical structures -- truly, it is the case that the eyes do not see what the mind does not know
- LGN also receives input from brainstem nuclei involved in attentiveness, which may be another way that attention modifies not only how well we see but what we see



Histologically, the LGN is a 6-layered structure

- Neurons from ipsilateral eye terminate on layers, 2, 3, and 5

- Neurons from contralateral eye terminate on layers, 1, 4, 6

- Layers 1 and 2 contain larger nuclei than those of layers 3-6 and are called magnocellular and parvocellular, and they contain input form the magnocellular (bigger receptive field, fast processing, low-contrast stimuli -- important for motion) and parvocellular (smaller receptive field, important for color, detail, and shape) retinal ganglion neurons, respectively. 

- the non-staining layers in between are koniocellular (greek for dust) - unknown function 

 LGN on MRI

now why does this GBM patient have homonymous hemianopsia? it looks like the T2 signal is invading into the distal optic tract, possibly LGN, and all of the bundles of right optic radiation. Remember: when its GBM, the T2 is not edema, its tumor.

Other places where retinal nuclei project:

- superior colliculus - important for coordinating eye movements - fixation (staring at one point, adjustments only to compensate for head motion), smooth pursuit (following one object smoothly over time), saccades (rapid movement in one direction) and vergence (eyes moving independently to maintain binocular/focused motion). Outputs go, as one would expect, to the nuclei of CN 3, 4, and 6. In higher primates, most of input to superior colliculus is not retina but visual cortex and frontal eye fields. 

- suprachiasmatic nucleus in hypothalamus - important for circadian rhythms and day/night cycles 

- pretectum - pupillary light reflex 

LGN to Visual cortex 

- There are two or three tracts depending on who you asked, and they can be named upper/lower, anterior/middle/posterior, ventral/dorsal etc. But the point is there is a bundle of fibers that carry upper visual quadrant information (meyer's loop) through temporal lobe and that terminates in the lower bank of the calcarine sulcus (lingual gyrus) and a bundle of fibers that carry lower visual quadrant information through the parietal lobe that terminates in the upper bank of the calcarine sulcus (the cuneus)
- The anatomy of meyer's loop is highly variable -- in some people, it will project all the way to the temporal pole. The traditional teaching that anterior/inferior temporal lobe is safe to resect is not always 100% true.

V1 lines the calcarine sulcus and extends into occipital pole

From V1, there is the dorsal stream "where" stream into parietal lobe (identifying motion, and relative positions in space of objets) and the ventral stream "what" stream into temporal lobe (object and person recognition). More on that next time 

sources

http://journal.frontiersin.org/article/10.3389/fnana.2010.00015/full

http://thebrain.mcgill.ca/flash/i/i_02/i_02_cr/i_02_cr_vis/i_02_cr_vis.html

http://iovs.arvojournals.org/article.aspx?articleid=2480802

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3236807/pdf/nihms319970.pdf