Medical Imaging 101 pt 2: CT

Medical Imaging 101 pt 2: CT

Following last weeks ScienceSunday post (http://goo.gl/pbeiY) I will focus on x-ray computed tomography (CT) today. At it’s basic level, it has an x-ray source and detector, that are on a gantry that rotates around the patient. CT scanners can have x-ray beams that are parallel, fan-beams, or cone-beams. Most of the clinical CT scanners today are helical or spiral. They have multiple rows of detectors (called slices). The x-ray source moves in a helical trajectory because the patient bed is moving while the x-ray source is rotating. This method provides higher resolution and high speed. A complete image can be obtained in less than a second. Here’s an excellent slide show http://goo.gl/Y0zTt, it’s a bit long though. For a look under the hood check out this previous #ScienceSunday post http://goo.gl/kdrcA

Physics of the raw data acquisition

So what is an x-ray? An x-ray is a form of electromagnetic radiation. Clinical CT machines use x-rays with energy in the 30 – 100 keV range, generated by an x-ray tube. An x-ray tube is basically a vacuum tube with a cathode, an anode, and a target (metal). The electron is accelerated in the vacuum and hits the target and produces x-ray photons (usually orthogonal to the direction of the electron beam). The effect is called bremsstrahlung. 

The x-ray photons can detected via semiconductor detectors which convert the x-ray photons to electron-hole pairs. A lot of modern CT detectors are flat panel detectors where a scintillator (a substance that takes ionizing radiation and produces luminescence) screen converts the x-ray photons into the visible range. Then a standard CCD (like your digital camera) can be used.

Contrast

Unlike other imaging modalities, CT does not “detect” anything endogenous. Therefore there is a lot of background, including the patient table. The density of the object determines how many x-ray photons hit the detector. A dense object like bone will absorb a lot of photons and have a strong “signal” in the image. Foam is not dense, and is often used to position a patient’s head or limb. In the image, the foam will barely be visible. External contrast agents can be used to enhance the image (contrast). Iodinated compounds are giving intravenous (IV) to visualize blood vessels and organs. Barium can be given orally or as an enema to visualize the GI tract. For research, gold nanoparticles are being used either IV or intraperitoneal (IP, in the abdominal cavity). The first image below demonstrates the benefit of injecting a contrast agent.

How is the 3D image made

2D projections are acquired as the x-ray source is rotated around the patient. The projections are essentially a radiograph like you might get if you go to the ER with a broken bone. The computed tomography part of x-ray CT is where you use computer algorithms to convert the 2D projections into a 3D image. CT uses Filtered Back Projection (FBP). The early CT scanners used parallel beams which is easier to describe FBP. In that case, each voxel is like a weighted sum of the cosine-corrected, filtered projection value of the x-ray attenuation. The tutorial linked above and the second image below demonstrate how the projections are added to generate the image. The video below demonstrates a series of 2D projections. The more projections you acquire the better the image becomes, to a point. The video below is of the Saguinus monkey skull from a previous post here: http://goo.gl/kZQAV

Strengths

CT’s strength is the relatively high spatial resolution in all directions, i.e., isotropic. The speed of acquisition is another strength. Also, CT can image without any contrast agent. Lung imaging is superb because of the natural contrast between air and tissue.

Weaknesses

The major weakness is radiation. For repeated imaging, e.g., to track response to therapy, the patient would receive repeated doses of radiation. For children this is even more of a concern. However, the new multislice CTs often have lower dose than the older single slice scanners. The other weakness is relatively poor sensitivity. With small animal imaging, the contrast difference can be limited even when using a contrast agent (see the first image below).

Something unique

One interesting ability with CT is the ability to treat and image. In radiation oncology, a linear accelerator or linac is used to image and treat cancer patients. You can check out the Wiki http://goo.gl/Jp8kL.

#ScienceSunday   #ScienceEveryday  curated by Allison Sekuler Rajini Rao Robby Bowles and me.

#CHMedicalImagingSeries

4 thoughts on “Medical Imaging 101 pt 2: CT”

  1. Nice intro. I work with detection and characterization of quantum light sources. We use a homodyne detector to measure the noise fluctuations of the light along different directions in the so-called phase space of phase and amplitude. From these complementary slices we can reconstruct the full quantum state of the light. This method is conceptually and computationally very similar to medical tomography, so it was called quantum state tomography.

    Interesting to read a bit more about the original version  🙂

  2. Jonas Neergaard-Nielsen look forward to hearing more about Quantum state tomography. Be sure to tag #ScienceSunday  if you do post about it.

Leave a Reply