SPECT is, in many ways, very much like a CT scan in radiology
The differences are:
The radiation source is emission vs. CT is transmission
Nuclear medicine looks at physiology, not anatomy
SPECT resolution has a matrix size of 64 or 128, whereas CT is 512 or 1024
Acquisition time in CT is VERY fast, but SPECT is VERY slow
In helical CT, the x-ray tube spins around the patient as the table moves. This allows us to image a significant part of the human anatomy.
SPECT's FOV is limited to the size of the detector
The computer algorithm or image reconstruction that produces tomographic slices is similar in both systems
The above image shows how a detector or a set of detectors collects information within a 360 radius around the organ of interest
The acquired data is then reconstructed via a filtered backprojection protocol, where the images are projected back at each angle acquired
Most SPECT systems have two detectors, however, three detectors systems are available
The following acquisition parameters are used and will vary pending the type of procedure being acquired:
One-hundred-eighty degree rotation, cardiac only
Usually, cardiac is imaged at a 64 matrix with 32 stops
Most 360-degree acquisitions are set at a matrix 64 with 64 stops
In certain procedures, a 128 matrix may be applied, and usually 128 stops are acquired
As a general rule, acquisition time should be no more than 30-minutes
Once the data is collected, the raw data can be displayed in a cine format, as seen above
The above acquisition is of a pediatric brain scan
Raw data is the reconstructed/processed that will produce the tomographic slices
Filtered backprojection is applied, and imaging filters are used to enhance image quality
The goal of filtering is to remove background counts and reduce noise
Can you recall the definition of image noise?
The goal is to only keep the true counts
Transverse, sagittal, and coronal images are generated and "centered," as seen in the displayed above
Again, it is essential to realize that the concept presented today is fundamental
From the above rotating SPECT images, the three types of slices are displayed (see above image)
It should be noted that these slices are only 2 pixels thick and have approximately 15 slices per each angle that are not being displayed (transverse, sagittal, and coronal)
Here is another example of ALL the slices you would see in a normal brain scan
From the acquired data, the lack of uptake in the frontal lobe of the brain is demonstrated
Moya Moya disease may starts with pediatric patients where calcification within the arterial structure of the brain occurs. The result is a loss of blood flow to specific areas of the brain
Moya Moya disease - in Japanese, it means "puff of smoke." Refer to the link
Acquisition settings to consider when acquiring a SPECT Bone Scan
Collimation - HR or UHR
Energy Setting - 140 keV with 20% window
Rotation - 360 elliptical or body contour
Matrix 64 with 64 slices
Matrix 126 with 128 slices
Two heads are better than one
Time per slice calculation - In general, you should never image longer than 30 minutes and a 2-headed SPECT system is preferred
Calculate the time per stop on a 30-minute acquisition - see above
Other points of interest
Why does SPECT have fewer counts than a spot view?
Even though a SPECT scan may have more total counts, the resolution is based on the number of counts acquired in a single stop
Is image shows the position of transverse, sagittal, and coronal angles
Here is an example of where a spot view on a bone scan seems normal, however, when SPECT images are taken abnormal uptake is seen