ACR Ghosting (and Uniformity)

[YassineAzma]

• Adds an ACR ghosting script and associated unit tests that calculates the percent signal ghosting of slice 7 of the ACR phantom. This is done within the new HazenTask framework. The methodology is as follows:
  • Calculates the centroid of the object following morphological operations and a centre-of-mass calculation.
  • This step also produces a binary mask.
  • Define a large ROI, of 200cm² area and offset from the top of the phantom.
  • Elliptical ROIs of 10cm² area are then drawn between the edges of the FoV and the phantom along the cardinal directions. These ROIs are stretched along the long-axis of the ellipses depending on how close they are to the FoV edge and phantom edge. The 10cm² area of each ellipse is maintained.
  • The mean values from each ROI are used to calculate the percent-signal ghosting. Which is defined as: $$PSG=100 {|(north + south) - (west + east)|\over(2large ROI)} $$
  • These results are visualised in the report, as shown below:

Additional edits

• Refactored ACR uniformity script and associated unit tests to work within the HazenTask class.
• Associated modifications to init.py to (re)add the necessary hazen CLI functions for the ACR uniformity and ghosting scripts.

N.B: There is no existing issue for the ACR ghosting test.

[github-actions[bot]]

Coverage Report

File	Stmts	Miss	Cover	Missing
hazenlib
HazenTask.py	24	3	88%	29–31
__init__.py	147	38	74%	135, 139, 149, 154, 191, 198–203, 214, 217–224, 244–246, 264–266, 285–287, 296, 301, 307, 358, 378–380, 388–389, 393
relaxometry.py	317	90	72%	238–256, 630, 689–691, 745, 793–815, 833–848, 1173–1176, 1185–1188, 1200–1213, 1216–1221, 1233–1263
shapes.py	20	9	55%	13, 16, 24–29, 32
snr_map.py	108	5	95%	403, 408–410, 439
tools.py	82	6	93%	43–50, 89
hazenlib/tasks
ghosting.py	150	51	66%	18–32, 47, 109–110, 114, 124–125, 151–153, 170–172, 218–256
relaxometry.py	7	2	71%	10–11
slice_position.py	117	21	82%	31, 40–41, 103–104, 130, 210, 217–233
slice_width.py	356	52	85%	34–37, 107, 166–186, 451, 456–457, 463, 468, 530–531, 780–821
snr.py	163	46	72%	62–67, 221–231, 273–277, 314–327, 332–340, 369–382
snr_map.py	7	7	0%	1–11
spatial_resolution.py	294	90	69%	36–39, 63–121, 134, 219, 255–262, 287, 413–449
task_acr_ghosting.py	116	42	64%	33–53, 91–93, 123–125, 161–194
task_acr_uniformity.py	89	32	64%	34–54, 121–138
uniformity.py	78	19	76%	42–45, 91–92, 99, 133–147
TOTAL	2112	513	76%

Tests	Skipped	Failures	Errors	Time
180	0 :zzz:	0 :x:	0 :fire:	2m 24s :stopwatch:

[github-actions[bot]]

Coverage Report

File	Stmts	Miss	Cover	Missing
hazenlib
HazenTask.py	24	3	88%	29–31
__init__.py	155	44	72%	136, 140, 150, 155, 192, 199–204, 215, 218–225, 245–247, 265–267, 286–288, 297, 302, 308, 359, 379–381, 384–386, 388–390, 398–399, 403
relaxometry.py	317	90	72%	238–256, 630, 689–691, 745, 793–815, 833–848, 1173–1176, 1185–1188, 1200–1213, 1216–1221, 1233–1263
shapes.py	20	9	55%	13, 16, 24–29, 32
snr_map.py	108	5	95%	403, 408–410, 439
tools.py	82	6	93%	43–50, 89
hazenlib/tasks
ghosting.py	150	51	66%	18–32, 47, 109–110, 114, 124–125, 151–153, 170–172, 218–256
relaxometry.py	7	2	71%	10–11
slice_position.py	117	21	82%	31, 40–41, 103–104, 130, 210, 217–233
slice_width.py	356	52	85%	34–37, 107, 166–186, 451, 456–457, 463, 468, 530–531, 780–821
snr.py	163	46	72%	62–67, 221–231, 273–277, 314–327, 332–340, 369–382
snr_map.py	7	7	0%	1–11
spatial_resolution.py	294	90	69%	36–39, 63–121, 134, 219, 255–262, 287, 413–449
task_acr_ghosting.py	116	42	64%	31–51, 89–91, 121–123, 159–192
task_acr_uniformity.py	89	32	64%	16–36, 103–120
uniformity.py	78	19	76%	42–45, 91–92, 99, 133–147
TOTAL	2120	519	76%

Tests	Skipped	Failures	Errors	Time
180	0 :zzz:	0 :x:	0 :fire:	2m 4s :stopwatch:

[YassineAzma]

@Lucrezia-Cester Do you know what this acr_snr CLI function that keeps failing is? I didn't include that, so I have no idea.

Edit: I've went through Laurence's pull request for the HazenTask refactoring and it seems to be what he renamed the acr_uniformity task to. I'll delete it.

[Lucrezia-Cester]

Hi @YassineRMH thank you so much for fixing that, I'll check the uniformity script and then approve this PR soon

[Lucrezia-Cester]

Hi @YassineRMH, by the way, feel free to add issues, and in general it would be best to separate different tasks in different PRs, in this case, ghosting from uniformity

[YassineAzma]

Hi @YassineRMH, by the way, feel free to add issues, and in general it would be best to separate different tasks in different PRs, in this case, ghosting from uniformity

Will do.

[Lucrezia-Cester]

Hi @YassineRMH, before I merge, would it be possible for you to add the docstirng for the unifromity script the same way you did for the ghosting? I am keen to keep track of sources.

[YassineAzma]

@Lucrezia-Cester Potentially stupid question, but what do you mean by docstring?

Edit: Done.

[Lucrezia-Cester]

Hi @YassineRMH, is there a link to the ACR guidance?

[YassineAzma]

@Lucrezia-Cester Do you want me to include a link to the guidance in my docstrings?

If not, it's here:

https://www.acraccreditation.org/-/media/acraccreditation/documents/mri/largephantomguidance.pdf

[Lucrezia-Cester]

Hi @YassineRMH, i just added it, I'll merge now

[YassineAzma]

@Lucrezia-Cester One last thing, I was trying to edit this before you merged:

On line 6 in the task_acr_ghosting.py script, could you please change uniformity to percent-signal ghosting. I can also make a separate pull request for this, but it seems very overkill.

[Lucrezia-Cester]

Hi @YassineRMH, apologies for not waiting for this edit. Will change it myself.

[YassineAzma]

@Lucrezia-Cester Thanks for all your help. 👍