Variable Lung Density Consideration in Attenuation Correction of Whole-Body PET/MRI

Harry R. Marshall; Frank S. Prato; Lela Deans; Jean Théberge; R. Terry Thompson; Robert Z. Stodilka

doi:10.2967/jnumed.111.098350

Article Figures & Data

Figures

Tables
Additional Files

Download figure
Open in new tab
Download powerpoint
FIGURE 1.
Relationship between MRI_lungSTE signal and CT number. Points on scatterplots represent voxels (A), coronal slices (B), and all lung tissue (C). Line-of-best-fit equations are y = 1,116x − 916, y = 1,223x − 932, and y = 1,251x − 937, respectively. Data from all 5 canines are plotted on each scatterplot. Each point’s color represents canine’s respiratory state, where dark blue is FRC, red is PIP = 8 cm of H₂O, and light blue is PIP = 16 cm of H₂O. As expected, increased volume of air in lungs reduces both MRI and CT signals.
Download figure
Open in new tab
Download powerpoint
FIGURE 2.
Axial slice through 4 pre-μ-maps obtained from canine 5 at PIP = 16 cm of H₂O. (A) CT-based image. (B–D) MRI-based images incorporating information from MRI_lungSTE via global (B), slice-by-slice (C), and voxel-by-voxel mappings (D). Window and level are identical for all images and set to emphasize lungs.
Download figure
Open in new tab
Download powerpoint
FIGURE 3.
Three metrics of PET_MRI’s quantitative fidelity broken down by tissue type and μ-map class. Metrics include (A), R² (B), and E (C). Leftmost blocks of data are from within lungs, middle blocks from soft tissue, and rightmost blocks from bone. Each point within given block corresponds to different PET_MRI reconstruction. Error bars indicate 95% confidence intervals.
Download figure
Open in new tab
Download powerpoint
FIGURE 4.
Sample profiles through PET reconstructions of canine at FRC. (A) Transaxial slice through CT-based μ-map for anatomic reference. Profiles were taken through cyan line that can be seen to pass through humeri, lungs, and heart. (B) Corresponding slice through PET_CT image. (C) Profiles through PET₋₉₀₀ and PET₋₄₀₀. Profiles through all other PET_MRI images are bounded by these curves. Solid line represents truth as obtained via PET_CT. (D) Profiles through PET_global, PET_slices, and PET_voxels. Again, black line represents truth.

View popup

TABLE 1

Results of ANOVAs on Canine PET_MRI Metrics of Overall Quantitative Fidelity and Quantitation Error in Myocardium

Metric	Factor	F	P
	Respiratory state	4.45	0.01
	Tissue	271.41	<0.0001
	μ-map class	11.45	<0.0001
R²	Respiratory state	13.72	<0.0001
	Tissue	899.97	<0.0001
	μ-map class	3.10	0.0001
E	Respiratory state	0.13	0.88
	Tissue	570.71	<0.0001
	μ-map class	9.35	<0.0001
Magnitude of relative error in VOIs	Respiratory state	44.65	<0.0001
	VOI	109.21	<0.0001
	μ-map class	25.63	<0.0001

View popup

TABLE 2

Average Magnitude of Relative Error in VOIs as Function of μ-Map

	HU
VOI	−900	−850	−800	−750	−700	−650	−600	−550	−500	−450	−400	Global	Slices	Voxels
Right lung	30 (10)	23 (11)	17 (11)	13 (9)	12 (7)	13 (9)	17 (13)	23 (17)	31 (19)	40 (21)	50 (22)	13 (11)	10 (8)	5 (3)
Left lung	34 (13)	28 (14)	22 (14)	18 (12)	15 (10)	14 (10)	16 (12)	20 (14)	26 (18)	33 (22)	41 (24)	8 (5)	8 (5)	4 (4)
Peripheral LV	17 (7)	14 (7)	11 (6)	9 (6)	6 (5)	5 (4)	4 (3)	5 (4)	7 (6)	9 (7)	12 (8)	4 (3)	4 (4)	5 (4)
Central LV	7 (2)	6 (2)	5 (2)	4 (2)	3 (2)	2 (1)	2 (2)	2 (2)	2 (2)	3 (3)	4 (3)	1 (1)	2 (1)	2 (1)
Vena cava	18 (11)	15 (8)	12 (7)	9 (8)	9 (9)	12 (9)	17 (11)	22 (13)	28 (14)	35 (15)	42 (16)	15 (9)	14 (8)	23 (6)
Vertebral body	16 (3)	15 (3)	15 (3)	14 (3)	13 (3)	13 (3)	12 (3)	11 (4)	11 (4)	10 (4)	10 (5)	12 (3)	12 (4)	11 (4)
Dome of liver	6 (5)	6 (4)	6 (4)	5 (4)	5 (4)	5 (5)	5 (5)	6 (6)	6 (6)	7 (7)	8 (7)	5 (5)	4 (5)	5 (5)
Chest wall	5 (4)	5 (4)	6 (4)	7 (4)	7 (5)	8 (6)	9 (7)	10 (8)	11 (9)	12 (10)	14 (10)	9 (6)	9 (6)	7 (5)

LV = left ventricle.
Data are mean error recorded as percentage of true activity, with SD in parentheses. Best result in each VOI is in boldface type, whereas results that are not statistically different from best result according to Tukey test are italicized.