Dicom data obtained from the Vanderbilt MUGA phantom

These data may be helpful to anyone:-

 

First there is a set of 3 images with background included

 

Because background correction is a major source of error the next set are free of background,

you can easily process these manually to confirm the “true” values. The way the phantom works

these “true” values apply to both sets of images (with/without background)

 

Finally there are a set of some serial acquisitions in zip files (with background).

In each zip file the acquisitions are all similar, e.g. they have the same “true” EF and similar count density.

The only thing that changes much in each zip are the random statistical variations.

There are also minor changes in count density due to radioactive decay.

NB one file from each high count zip is identical to the single file examples in the links higher up.

 

Thoughts on Phantom data. Of course this data will not solve all you MUGA consistency issues.

I think the data from this phantom is rather too easy for the automatic software to process.

Too easy for the manual operator too for that matter.

So in other words even if you are getting great consistency with this data, it does not mean you are

getting the same consistency with patient data. But if you get variations between operators or software

on this data it could be useful for analysing what is going wrong.

On the plus side, at least compared with any patient data, this phantom data has verifiable “true” results.

 

If anyone wants to compare results, request some different data etc I would be happy to hear from you.

My personal email address is below or you can contact me on the med phys listserver.

Mike Avison

------------------------------------------------

The “True” LVEF values are:-

Low: 30%

Med: 51%

High: 75%

 

------------------------------------------------

Images with background structures present

Surname = MUGA,  First name = Test, ID = Vanderbilt

Description: Description: Description: Description: Description: Description: Description: Description: Description: MUGA_with_bg

High EF

Medium EF

Low EF

 

------------------------------------------------

 

Images without background structures present

Surname = MUGA,  First name = Test Thursday

Description: Description: Description: Description: Description: Description: Description: Description: Description: MUGA_without_bg

High EF

Medium EF

Low EF

 

-------------------------------------------------

ZIP files of multiple studies with background

High Count Studies – summed LV counts vary from 400 to 660k

Eleven Studies with High EF and High counts

Thirteen studies with Med EF and High counts

Thirteen studies with Low EF and High Counts

 

Low Count Studies – summed LV counts vary from 200 to 330k

Eleven studies with High EF and Low counts

Twelve Studies with Med EF and Low counts

Thirteen studies with Low EF and Low Counts

-------------------------------------------------

Acquisition parameters

Collimators: LEGP

Acquisition time: 1000s for hi count and 500s for lo count

Activity concentration: 0.3MBq/ml

Activity in Left Ventricle: 12MBq

Activity in Background (when present): 80MBq

Time bins: 16

Camera: Philips Brightview XCT

Zoom factor: 2.18

Field of view: 273 x 273 mm

Pixel size: 4.23mm

Matrix: 64x64

Heart rate: 60bpm