SPECT imaging
Single photon emission computed tomography (SPECT) is the method of obtaining cross-sectional nuclear images (similar to CT in x-ray imaging).
- Single photon:
- SPECT uses single gamma photon detection that are produced by gamma photon decay
- c.f. PET which uses the simultaneous detection of the two gamma photons that arise from positron decay
- Emission:
- Radioactivity used to create image is emitted from patient rather than transmitted through patient from an outside source as is done in x-ray imaging
- Computed tomography:
- Slices are imaged that can be reconstructed into 3D data
SPECT can be used to image any radiopharmaceutical in which:
- The distribution does not change significantly during the image acquisition time (20-40 minutes)
- Acquisition time long enough for sufficient amount of gamma photons to be collected
Equipment
Camera / detector
a. Single head gamma camera
- Rotated around the patient during image acquisition
- Long image acquisition times
- No longer commonly used
b. Multiple head gamma camera
- Dual head, large field of view camera
- Housed on a gantry with slip ring technology that can rotate the cameras around the patient
- The cameras can be positioned in either an H-configuration or an L-configuration relative to each other
Hybrid SPECT/CT
- X-ray source and x-ray detector array placed between the gamma camera heads
- Anatomical CT and functional SPECT images then fused
- CT information can also be used to correct for attenuation in the SPECT images
Gantry
Needs to have:
- Accurately aligned centre of rotation
- Constant rotational speed
- Detectors aligned parallel to axis of rotation
Collimator
Important to use high resolution collimator
- Maximise spatial resolution throughout depth of the patient
- Reduce image distortion during reconstruction
Hole direction
- Parallel holes
- Hole and septae size as uniform as possible
- Non-parallel holes
- Can only be used with circular (i.e. not body contouring) orbits e.g. heads
Collimator type
- Fan beam collimator
- Used for brain imaging
- Utilises magnification - uses more of the detector field of view to collect the image data
Ensure smallest camera-patient distance but maintain safe distance
- Infra-red beams fitted to collimator face that enable automatic body contouring to minimise the detector-patient distance and optimise image quality
- Fitted with pressure sensitive safety devices to prevent any contact between the collimators and the patient
Patient table
- Needs to be comfortable due to long image acquisition times
- Low attenuation of gamma photons to allow photons to pass through and enable 360 degree acquisition
Image acquisition
Matrix
- Determines maximum resolution and image noise (counts per pixel)
- Modern dual headed system = 128 x 128 matrix
- To reduce image noise (at expense of resolution):
- Increase slice thickness
- Smoother reconstruction filters
- Display slice data in 64 x 64 pixel matrix
Views
- 20-40 sec per projection frame
- Heads rotated in continuous or 'step and shoot' mode
Minimising artefacts
- Minimise patient movement
- Injection site (very high count density) should be kept out of the field of view
- Arms above heads for chest and abdominal imaging to remove radiation attenuation and minimise patient-detector distance
Specific scans
- Cardiac SPECT
- Heart is located off centre
- Imaged over 180 degrees from LPO to RAO projection with heads in L mode to reduce detector distance and attenuation in tissues
- ECG gating used to demonstrate cardiac wall motion (8 frames per cardiac cycle acquired into 64 x 64 matrix for each projection angle)
Out now in paperback and on Kindle
Written by radiologists, for radiologists with plenty of easy-to-follow diagrams to explain complicated concepts. An excellent resource for radiology physics revision.
Note: Not available in all regions.
Reconstruction
1. Creation of 1D profiles from each projection angle
- As the camera rotates around, it creates a 1 dimensional view of the measured radioactivity for each angle.
2. Filtering of profiles
- Compromise between noise reduction (degree of smoothing) and preservation of image (resolution)
- Smoothing usually defined by cut-off or critical frequency of the filter (maximum spatial frequency present in image)
3. Processing of data to create reconstructed slice image
- Back projection or iterative reconstruction of the filtered profiles create the reconstructed slice image. More detailed accounts of backprojection and iterative reconstruction is available here.
Correcting for attenuation:
- Analytical method: apply algorithm that assumes uniform attenuation and then adjusts pixel counts depending on distance from camera and edge of patient
- Direct measurement: CT data can be used to calculate an attenuation map and adjust the pixel counts according to this
