deletions | additions       diff --git a/I_asked_colleagues_that_work__.tex b/I_asked_colleagues_that_work__.tex  new file mode 100644  index 0000000..c05b9c3  --- /dev/null  +++ b/I_asked_colleagues_that_work__.tex 
...
I asked colleagues that work on ADNI data at UCSF, and they verified that the same healthy controls were not scanned at multiple sites, and so I wouldn't be able to calculate a $CV_a$ from that data. However, its still really important to compare our results to other harmonization efforts, so I have added 2 tables that do this. In the first, our between site ICC results pre- and post- calibration are compared to the between site ICC results of \cite{cannon2014}. In \cite{cannon2014}, the sites were harmonized and an ADNI phantom was used to correct gradient distortions. The authors ran FSL's FIRST for subcortical segmentation, and a cortical pattern matching algorithm for gray/white segmentation. They then calculated the between-site ICC using variance components in the same way we did. The table below was added to the manuscript:  \begin{table}   \begin{tabular}{ c c c c }  ROI & ICC BW & ICC BW Cal & \cite{cannon2014} ICC BW \\   \midrule  GMV & .78 & .96 & .854 \\   WMV & .96 & .98 & .774 \\   Thal & .61 & .73 & .95 \\   Hipp & .75 & .84 & .79 \\   Amyg & .56 & .74 & .76 \\   Caud & .82 & .91 & .92 \\   \bottomrule  \end{tabular}   \caption{Between-site ICC comparison to the study by \cite{cannon2014}, where MRI sequences were standardized and subcortical segmentation was performed using FIRST, and cortical segmentation using cortical pattern matching. ICC BW and ICC BW Cal were calculated using our multisite healthy control data, where ICC BW Cal was calculated as the between site ICC of volumes after applying the scaling factor from a leave-one-out calibration. Other than the thalamus (Thal), we found that the between-site ICCs were comparable to \cite{cannon2014} for the amygdala (Amyg), caudate (Caud), and even higher for the hippocampus (Hipp), gray matter volume (GMV) and white matter volume (WMV)}   \label{comparetocannon}  \end{table}  The only ROI that does not compare to \cite{cannon2014} is the Thalamus. It is possible that the FIRST algorithm is more reliable at segmenting this structure. I included another table by \cite{jovicich2013brain} where again, sites were harmonized, different control subjects were scanned at each site, but the authors used the freesurfer cross-sectional algorithm that we used. Instead of calculating between-site ICC's, they calculated the average within-site ICCs for each ROI. The following table (which is now included in the manuscript) compares our within-site ICC's pre- and post- calibraiton to \cite{jovicich2013brain} average within-site ICC values:  \begin{table}   \begin{tabular}{ c c c c }  ROI & ICC WI & ICC WI Cal & \cite{jovicich2013brain} ICC WI Average \\   \midrule  LV & 1 & 1 & $.998 \pm 0.002$ \\   Thal & .86 & .84 & $0.765 \pm .183$ \\   Hipp & .93 & .93 & $0.878 \pm .132$ \\   Amyg & .89 & .86 & $0.761 \pm .134$ \\   Caud & .97 & .97 & $0.909 \pm 0.092$ \\   \bottomrule  \end{tabular}   \caption{Comparing the within-site ICC before and after leave-one-out scaling factor calibration with the cross-sectional freesurfer results of \cite{jovicich2013brain}, where scanners were standardized, and the average within-site ICC is shown. The within-site ICCs of our study fall within the range of \cite{jovicich2013brain}, which shows the that sites in this study are as reliable as those in \cite{jovicich2013brain}.}   \end{table}  Here, we see that the within-site thalamus ICC values fall within the range of \cite{jovicich2013brain}, along with the other ROIs.