Naked Science Forum

Life Sciences => Physiology & Medicine => Topic started by: vhfpmr on 12/02/2021 14:53:37

Title: CT Scanner Operation
Post by: vhfpmr on 12/02/2021 14:53:37
Is it possible to get a CT scan of the head with the patient table stationary, and not traversing through the scanner? How long does a head scan typically take.
Title: Re: CT Scanner Operation
Post by: evan_au on 13/02/2021 00:10:59
If you just wanted a single slice through the head, I guess you could do that
- Although I understand that the beam shape is somewhat fan-shaped
- And the imaging algorithms are probably optimized to deal with a helical scan (moving patient)

But why would you want just a single slice?
- Most medical problems exist in 3 dimensions
- The exact location of medical problem is somewhat uncertain, so aligning this 2D slice is difficult
- And most doctors are 3-dimensional, so they will disrupt things around this 2D slice.
- So imaging a 3D volume has real advantages for the doctor
- Even if it means a slightly higher X-Ray dose for the patient
Title: Re: CT Scanner Operation
Post by: alancalverd on 13/02/2021 11:36:20
A multislice CT scanner can take up to 64 parallel slices in a single rotation of the tube head (or more for some cardiac research units) so in principle can cover a small 3D volume without moving the patient. If you want really high spatial resolution it's a good idea to shift by half a slice width and scan again, then interpolate the images to "smooth over the joins". Scan time can be down to tens of milliseconds for a conventional system or microseconds for a single slice "electron beam CT."

There are clinical advantages in MRI scanning to have the patient standing or sitting, and I've owned a few such systems, but I haven't persuaded anyone of the diagnostic value of upright CT: a pity, because,  like MRI, my design involves a stationary patient and a moving scan field.

You can also use "cone beam" CT  for small volumes. These systems are very popular in dentistry, essentially taking several hundred complete images up to 10 x 10 cm  in a single 10 second rotation and interpolating the reconstruction. Resolution of around 2.5 lines per mm is standard but the software and x-ray spectrum are usually optimised for hard tissue (bones and teeth) as against medical CT which is optimised for resolving soft tissue inside a skull. 
Title: Re: CT Scanner Operation
Post by: Hayseed on 14/02/2021 00:48:39
I was scanned from my nipples down.  About 8 yrs ago.  I was stationary, the scanner moved.  UN-believable images.  Super detail.  I wish they would have done the whole body.  Too much metal in me now.
Title: Re: CT Scanner Operation
Post by: vhfpmr on 15/02/2021 14:12:08
A multislice CT scanner can take up to 64 parallel slices in a single rotation of the tube head (or more for some cardiac research units) so in principle can cover a small 3D volume without moving the patient. If you want really high spatial resolution it's a good idea to shift by half a slice width and scan again, then interpolate the images to "smooth over the joins". Scan time can be down to tens of milliseconds for a conventional system or microseconds for a single slice "electron beam CT."

There are clinical advantages in MRI scanning to have the patient standing or sitting, and I've owned a few such systems, but I haven't persuaded anyone of the diagnostic value of upright CT: a pity, because,  like MRI, my design involves a stationary patient and a moving scan field.

You can also use "cone beam" CT  for small volumes. These systems are very popular in dentistry, essentially taking several hundred complete images up to 10 x 10 cm  in a single 10 second rotation and interpolating the reconstruction. Resolution of around 2.5 lines per mm is standard but the software and x-ray spectrum are usually optimised for hard tissue (bones and teeth) as against medical CT which is optimised for resolving soft tissue inside a skull. 
Thanks AC.
So even with 64 slices, at 2.5 lpmm that would only cover 25mm. Presumably if you were looking for something like a brain haemorrhage you would need to scan most of the head? If you were doing only a partial scan I imagine there would be some time and care required to locate the correct place, quite apart from advance knowledge of that place itself. I'm intrigued by the idea of doing a slice in tens of milliseconds, if it were 50ms for example wouldn't that mean the head rotating at 1200rpm? I don't recall hearing a whine like a washing machine on spin cycle.

Presumably the NHS like horizontal scanners for the flexibility of scanning unconscious/incapacitated patients with the same machine?
Title: Re: CT Scanner Operation
Post by: alancalverd on 15/02/2021 15:22:57
 I quoted spatial resolution for a CBCT machine which in effect takes 180 or so complete planar digital x-rays with an inherent resolution of around 2.5 lp/mm and interpolates the data to provide a 3D image with a voxel length of around 0.2 mm.

There are indeed standard 64-slice cardiac CT scanners running at around 200 ms per rotation. OK, hundreds rather than tens!
Title: Re: CT Scanner Operation
Post by: vhfpmr on 15/02/2021 16:15:18
I've had a couple of abdominal scans where I got onto the table, was plumbed in for contrast, and then there was first a quick scan that appeared to be just for locating the datum for my position, followed by several slower scans. All in all it would have been a good 10-20 mins.

What interests me now is that I had a head scan last week to look for a brain haemorrhage in which I got onto the table which then slid into the scanner, then as I waited a few seconds for the scan to start the table withdrew again and the operator told me I was all clear. I doubt I was in the scanner for more than a minute.

The difference between the head scan and the abdominal scans is like chalk & cheese.
Title: Re: CT Scanner Operation
Post by: alancalverd on 15/02/2021 18:17:23
The difference between the head scan and the abdominal scans is like chalk & cheese.
Yes. We spend several years teaching radiographers the difference between an arse and an elbow - but they still get it wrong sometimes.

The abdominal scout scan is more often needed than for a head because the bits inside (to use the technical term) tend to float around between different patients, but the brain is nearly always located in the skull (apart from teenagers, where it is located in the pelvic region). It takes time for the contrast agent to reach wherever it is going and if you are looking for differential flow as a diagnostic, you may need a few scans  to measure it, whereas a brain hemorrhage will be fairly static so you either see it in a single shot or you don't.  I'm pleased they didn't.
Title: Re: CT Scanner Operation
Post by: vhfpmr on 15/02/2021 23:29:09
OK, thanks AC. I didn't realise brain haemorrhages are always big and conspicuous on a scan rather than difficult to spot.
Title: Re: CT Scanner Operation
Post by: alancalverd on 15/02/2021 23:38:58
By the time it is symptomatically obvious, it is usually pretty easily visible on a CT scan. Time for a deep breath. 
Title: Re: CT Scanner Operation
Post by: vhfpmr on 16/02/2021 15:31:17
If only it were that simple.
Title: Re: CT Scanner Operation
Post by: evan_au on 16/02/2021 21:13:19
Quote
alancalverd]The abdominal scout scan is more often needed than for a head
Speaking of abdominal scans, do they still use barium meals, or do they use MRI to image the digestive system?
Title: Re: CT Scanner Operation
Post by: alancalverd on 16/02/2021 23:29:52
A barium swallow or enema is quick, cheap, well tolerated, and generally easier to read than an MRI. In skilled hands it gives  information about flow and obstruction and the contrast of the barium coating means that there is little shadowing from the spine. 

MRI is still relatively rare, slow and expensive, and around 10 - 20% of patients cannot tolerate a conventional tunnel machine. Gastrointestinal MRI contrast agents are less effective than barium.

Erect studies are useful, but the only suitable MRI machines are even rarer, slower and more expensive (I used to sell them).