Table 3 SNR of (1.3, 1.3) ppm in 2D lipid composition spectra using DQF-COSY.

From: Phased-array combination of 2D MRS for lipid composition quantification in patients with breast cancer

 

Algorithms

Tumour specimens

(N = 17)

P-value

Healthy volunteers

(N = 15, 30 voxels)

P-value

Patients

Peritumoural adipose tissue

(N = 15)

P-value

Patients

Tumours in vivo

(N = 10)

P-value

External reference

Equal

59.5 (41.6–105.8)

 

469.5 ± 127.7

 

309.9 ± 112.5

 

37.8 (19.7–57.0)

 

Signal

75.4 (52.3–131.3)

 

621.4 ± 239.6

 

444.6 ± 215.9

 

42.9 (25.8–81.0)

 

S/N

76.9 (53.2–133.0)

 

637.4 ± 235.2

 

462.2 ± 219.2

 

46.1 (25.2–82.4)

 

S/N2

78.2 (53.3–133.4)

 

645.0 ± 229.1

 

472.4 ± 218.4

 

47.7 (24.8–81.9)

 

WSVD

82.3 (56.8–133.1)

0.246

825.6 ± 304.5

0.003

581.1 ± 291.6

0.022

50.5 (31.8–101.9)

0.241

nd-comb

82.4 (56.7–133.8)

0.049

812.1 ± 300.7

 < 0.001

570.7 ± 291.0

 < 0.001

48.5 (33.4–100.4)

0.575

AOC

82.5 (56.8–133.9)

0.003

825.7 ± 305.2

0.003

582.2 ± 293.2

0.100

50.2 (33.4–101.3)

0.575

Internal reference

Equal

39.5 (30.8–103.9)

 

457.4 ± 148.4

 

311.4 ± 154.6

 

37.4 (18.2–57.0)

 

Signal

54.1 (41.0–129.9)

 

609.8 ± 248.5

 

433.2 ± 230.8

 

41.2 (23.9–82.9)

 

S/N

57.5 (41.9–131.8)

 

624.9 ± 244.5

 

449.2 ± 237.2

 

43.6 (24.4–82.5)

 

S/N2

60.0 (41.6–132.2)

 

630.0 ± 240.4

 

482.5 ± 239.1

 

44.9 (24.6–81.4)

 

WSVD

81.9 (57.2–136.4)

833.6 ± 308.6

584.6 ± 294.7

49.8 (32.6–102.8)

nd-comb

64.2 (50.6–135.2)

 < 0.001

743.6 ± 319.8

0.001

543.6 ± 310.7

0.007

43.1 (27.5–99.3)

0.028

AOC

60.0 (46.1–130.8)

 < 0.001

781.4 ± 338.1

0.005

552.6 ± 312.9

0.034

46.0 (28.6–98.9)

0.059

  1. The median SNR and interquartile range are shown for non-normally distributed data while mean and standard deviation are shown for normally distributed data. P-value represents the comparison on SNR between the noise decorrelated algorithms using repeated measures ANOVA or Wilcoxon signed-rank tests. The SNR of the linear algorithms is also reported for reference purposes. Results are presented with WSVDi as a reference for comparison.
  2. AOC adaptively optimised combination, CV coefficient of variance, nd-comb noise decorrelated combination, WSVD whitened singular value decomposition.