The report tells you the diagnosis. The images tell you the anaesthetic — how much room is left, which way things are moving, and which of the safety facts on this page apply to your patient today.
Two rules get you started: anatomy is better seen on T1, pathology on T2 — because most pathology holds excess free water, which is dark on T1 and bright on T2. Everything else on this page is a safety fact rather than an interpretive one. A ferromagnetic aneurysm clip is an MRI contraindication and a titanium one is not; myelographic contrast lowers the seizure threshold; diagnostic cerebral angiography carries a 0.3% to 1.3% neurological complication rate. None of that is in the report.
A radiology report is written to answer somebody’s question, and usually that somebody is the surgeon. It will tell you what the lesion is. It will not reliably tell you the things that change what you do in the next hour: how much compensatory room is left inside the skull, which way the contents are moving, and whether this particular patient can safely be put in a magnet at all.
On T1, fat appears bright and water and CSF appear dark. On T2, fat is intermediate grey while water and CSF are bright. Cortical bone gives no signal at all, and flowing blood produces a signal void.
Because pathologic processes usually contain excess free water, they are dark on T1 and bright on T2. Hence the rule of thumb: anatomy is better seen on T1 and pathology on T2.
Contrast agents are mostly paramagnetic — gadolinium — and make vessels and lesions more conspicuous by increasing T1 signal. That is why post-contrast sequences are T1 weighted.
Three more sequences you will see named:
And CT remains the workhorse in the emergency setting, for the unglamorous reason that it is everywhere and it is fast — which makes it the modality for the agitated, unstable patient in whom a long scan is not an option.
This is the section to actually memorize, because none of it appears in a report.
MRI contraindications. Cardiac pacemakers, intraocular metallic fragments, cochlear implants, drug infusion pumps, neurostimulators for deep brain stimulation, and ferromagnetic aneurysm clips are contraindications to MR imaging. The danger is magnetically induced movement — dislodgement and injury to organs, and heating. Image degradation is the secondary concern.
But not all clips. Non-ferromagnetic aneurysm clips, of titanium alloy or pure titanium, have been shown to be MR compatible and safe. So the question is never “does this patient have a clip” but “what is the clip made of”, and the answer lives in an operation note that may be a decade old.
Occupational screening. Welders, grinders and metalworkers get screening plain radiographs of the orbits before MRI, to exclude a metallic foreign body.
Diagnostic cerebral angiography. A catheter is introduced, usually through the right femoral artery, and iodinated contrast injected for digital subtraction angiography. A four-vessel study covers both internal carotids and both vertebrals; a six-vessel study adds the external carotids. Mild intravenous sedation improves comfort and cooperation.
The risk figure worth carrying: neurological complications occur in 0.3% to 1.3%, of which 0.07% to 0.5% are permanent. Most complications are minor and transient — groin haematoma, femoral artery injury, minor allergic reaction — but cerebral infarction, seizure and death occur. Transradial access reduces vascular and bleeding complications and introduces its own: radial artery perforation and compartment syndrome.
Spinal angiography carries all of that plus spinal cord infarction from a spinal artery embolus. It should therefore follow demonstration of a vascular malformation by another modality, or be reserved for subarachnoid haemorrhage with a normal cerebral angiogram and a strongly suspected spinal source.
Myelography. Iodinated contrast is injected into the subarachnoid space, usually by lumbar puncture. The complication worth knowing is not the headache: contrast material lowers the seizure threshold. Patients with a seizure history are studied cautiously, and patients on drugs that lower the threshold — tricyclic antidepressants are the example given — are commonly instructed to stop three days before.
If you are shown a perfusion CT, the parameters are mean transit time, time to peak, cerebral blood volume and cerebral blood flow, related by CBF = CBV / MTT.
The normal grey matter values are worth holding because they are in the same units as the ischaemic thresholds from Neuro 1: CBV 4 to 6 mL/100 g and CBF 50 to 60 mL/100 g/min. When someone points at a region and says the flow there is 20, you now know what that means without asking.
Four features, and they take about fifteen seconds once you know them. The imaging features of intracranial mass effect are sulcal effacement, midline shift, basal cistern effacement, obstructive hydrocephalus — and herniation, which is tomorrow.
The degree of mass effect correlates with the level of consciousness. That sentence is the reason this is your business rather than the radiologist’s: the scan is telling you how much room is left before the drugs you are about to give change the volume of what is inside the skull.
No randomized trial addresses this.
The imaging physics is not in dispute. T1 and T2 behaviour, the nulling of CSF signal on FLAIR, the relationship between gadolinium and T1 signal, and the diffusion behaviour of cytotoxic against vasogenic oedema are all settled and reproducible.
The safety practice rests on a guidance document. The American College of Radiology's MR safe practices document is the standard reference for screening and for the device list, and it is guidance built on accumulated incident experience rather than on trials — because the events it exists to prevent are rare, catastrophic and unrandomizable.
The procedural risk figures are observational. The complication rate for diagnostic cerebral angiography comes from consecutive-series data rather than from a comparison, and it is the number worth knowing precisely because it is small enough to be dismissed and large enough to matter when the indication is soft.
Strength, in the guideline's own words The ACR document is styled a "guidance document on MR safe practices" and its recommendations are consensus positions of an expert panel.
The characteristic failure is not misreading an image. It is never looking at one — accepting a verbal summary of a scan as though it were the scan, and then being surprised in the room.
A report is written to answer a specific clinical question, usually the surgeon's. "Left frontal glioma with surrounding vasogenic oedema" is accurate and it does not tell you how far the midline has shifted, whether the basal cisterns are effaced, or whether the ventricles are dilated — which is to say it does not tell you how much room is left in the box before you give a drug that will change the volume of its contents.
Two habits close the gap and neither takes long. Open the images yourself before the case and look for the four features of mass effect: sulcal effacement, midline shift, basal cistern effacement, obstructive hydrocephalus. And when you cannot interpret something, ask — the radiologist and the surgeon have both already looked at it, and the question "how much shift is there" has a number as its answer.
There is a second, quieter version of this pitfall. The imaging safety facts on this page live in nobody's report. Whether the aneurysm clip in this patient's head is titanium or ferromagnetic determines whether they can have an MRI at all, and that information is in an old operation note rather than on the scan you are looking at.
Full text · Preston goes through LibKey to whatever Preston holds — the PDF where there is one — and works on a phone. The doi link always reaches the publisher's record, whether or not we hold a subscription. If you are reading on a computer, LibKey Nomad adds the same one-click PDF button to PubMed and publisher pages everywhere else; it is a desktop browser extension, so it does nothing on a phone or tablet.
A patient is having an MRI before craniotomy. Which statement about the sequences is correct?
A patient who had an intracranial aneurysm clipped twelve years ago is listed for an MRI. What determines whether the scan can proceed?
You open the CT of a patient listed for craniotomy. Which findings most directly change your anaesthetic plan?
No returning items before day 4. The retrieval window opens once there is something far enough back to be worth retrieving.