Erratum

Following publication of the original article [1] the authors contacted the team with the request that some additional information be added to the article. Please see below for details:

Additional Table: absolute variability

Variability (VAR) presented in the main article represents mean variability between test and retest, not absolute variability as stated. The authors would like to apologise for this error. The absolute variability is presented in the table below has been calculated in the following manner:

$$ \mathbf{Absolute}\ \mathbf{variability} = \left(\mathbf{2}*\left(\left|\mathbf{Test}\ \hbox{--}\ \mathbf{Retest}\right|\right)\right)\ /\ \left(\mathbf{Test} + \mathbf{Retest}\right)\Big)\ *\ \mathbf{100} $$
 

Raw SUV

SUVWB

SUVC

Frontal Lobe

6.05

1.01

4.36

(2.45)

(0.67)

(4.84)

Parietal Lobe

6.69

0.97

5.15

(2.33)

(0.77)

(5.10)

Temporal Lobe

5.92

0.61

3.78

(3.13)

(0.49)

(4.92)

Occipital Lobe

6.81

1.07

3.87

(2.83)

(0.88)

(3.79)

Hippocampus

5.82

1.57

5.35

(1.95)

(1.34)

(6.50)

Parahippocam-pal cortex

5.05

2.07

5.77

(3.78)

(2.58)

(7.96)

Posterior Cingulate

6.58

1.89

5.92

(2.60)

(1.33)

(5.97)

Amygdala

3.45

4.31

6.54

(4.50)

(3.17)

(8.56)

Cerebellum

7.18

4.18

 

(7.84)

(5.07)

Whole Brain

6.09

  

(2.73)

Mean (SD)

5.96

1.96

5.09

(3.41)

(1.81)

(5.95)

These analyses allow for easier comparison with other studies however do no alter the findings or discussion of the paper – SUV normalised to whole brain (SUVWB) shows the lowest variability in our cohort, with increased variability found in the unadjusted SUV and the SUV normalised to cerebellum. As compared to other test-retest studies discussed in the paper, the variability for SUV based methods remains favourable in this pilot cohort. These results should be validated in larger studies.