RepeatedMeasures
AG
2/10/2020
Steps:
- Filter in data from subjects which have completed MR1 and MR2
- Repeated Measures by Site
- Anova1 = Remove variance from scanner (Scale IQMs by Scanner) and complete repeated measures anova within subjects
- Anova2 = Remove indivdual variance across sessions (average T1 & T2) and complete anova for Scanner differences
1. Filter in data from subjects which have completed MR1 and MR2
Repeated_T1<-read.csv("Database.T1.clean.csv")
tmp1<-c("SubID","Session")
tmp1<-Repeated_T1[tmp1]
tmp1<-tmp1[duplicated(tmp1$SubID), ]
Completed_Subs<- c("SubID")
Completed_Subs<-tmp1[Completed_Subs]
df<-c("SubID", "Session" , "SITE", "Scanner", "MPRAGE_cnr", "MPRAGE_fwhm_avg", "MPRAGE_snr_total" ,"MPRAGE_snrd_total" )
rep<-Repeated_T1[df]
Completed.Repeated.df<-merge(rep,Completed_Subs, by="SubID" )
Completed.Repeated.df<-Completed.Repeated.df[complete.cases(Completed.Repeated.df),]
rm(Completed_Subs, tmp1, rep, Repeated_T1)2. Compute basic stat tables across scanner and site
## Step 2
Completed.Repeated.df$N<-1
ByScanner<-Completed.Repeated.df %>%
group_by(Scanner) %>%
summarise( ScannerCount=sum(N),
meanCNR=mean(MPRAGE_cnr), sdCNR=sd(MPRAGE_cnr),
meanFWHM=mean(MPRAGE_fwhm_avg), sdFWHM=sd(MPRAGE_fwhm_avg),
meanSNR=mean(MPRAGE_snr_total), sdSNR=sd(MPRAGE_snr_total),
meanSNRd=mean(MPRAGE_snrd_total), sdSNRd=sd(MPRAGE_snrd_total))
ByScanner## # A tibble: 4 x 10
## Scanner ScannerCount meanCNR sdCNR meanFWHM sdFWHM meanSNR sdSNR
## <fct> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 KU Sky… 112 1.61 0.166 4.04 0.139 10.4 0.504
## 2 NEU Pr… 60 1.83 0.262 4.14 0.145 10.5 0.523
## 3 PITT P… 118 1.70 0.200 4.15 0.149 10.1 0.508
## 4 PITT P… 2 1.49 0.0242 4.23 0.0522 9.67 0.373
## # … with 2 more variables: meanSNRd <dbl>, sdSNRd <dbl>BySITE<-Completed.Repeated.df %>%
group_by(SITE) %>%
summarise(ScannerCount=sum(N),
meanCNR=mean(MPRAGE_cnr), sdCNR=sd(MPRAGE_cnr),
meanFWHM=mean(MPRAGE_fwhm_avg), sdFWHM=sd(MPRAGE_fwhm_avg),
meanSNR=mean(MPRAGE_snr_total), sdSNR=sd(MPRAGE_snr_total),
meanSNRd=mean(MPRAGE_snrd_total), sdSNRd=sd(MPRAGE_snrd_total))
BySITE## # A tibble: 3 x 10
## SITE ScannerCount meanCNR sdCNR meanFWHM sdFWHM meanSNR sdSNR meanSNRd
## <fct> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 KU 112 1.61 0.166 4.04 0.139 10.4 0.504 65.2
## 2 NEU 60 1.83 0.262 4.14 0.145 10.5 0.523 58.7
## 3 PITT 120 1.70 0.200 4.15 0.148 10.1 0.508 52.1
## # … with 1 more variable: sdSNRd <dbl>NEU Repeated Measures MPRAGE Reports
DATA_NEU_Prisma<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="NEU Prisma",]
DATA_Kansas_Skyra<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="KU Skyra",]
DATA_PITT_Prisma_1<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="PITT Prisma 1",]
DATA_PITT_Prisma_2<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="PITT Prisma 2",]
DATA_PITT<-rbind(DATA_PITT_Prisma_2,DATA_PITT_Prisma_1 )
# Mann-Whitney U test (nonparametric test)
NU_MPRAGE_cnr.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_NEU_Prisma,
x = Session,
y = MPRAGE_cnr,
xlab = "Session",
ylab = "CNR",
type = "np",
conf.level = 0.99,
ggstatsplot.layer = FALSE,
messages = FALSE
)
NU_MPRAGE_cnr.RepeatedMs.plot
NU_MPRAGE_fwhm.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_NEU_Prisma,
x = Session,
y = MPRAGE_fwhm_avg,
xlab = "Session",
ylab = "FWHM",
type = "np",
conf.level = 0.99,
ggstatsplot.layer = FALSE,
messages = FALSE
)
NU_MPRAGE_fwhm.RepeatedMs.plot
NU_MPRAGE_snr.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_NEU_Prisma,
x = Session,
y = MPRAGE_snr_total,
xlab = "Session",
ylab = "SNR",
type = "np",
conf.level = 0.99,
ggstatsplot.layer = FALSE,
messages = FALSE
)
NU_MPRAGE_snr.RepeatedMs.plot
NU_MPRAGE_snrd.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_NEU_Prisma,
x = Session,
y = MPRAGE_snrd_total,
xlab = "Session",
ylab = "SNRd",
type = "np",
pairwise.annotation = "p-value",
outlier.tagging = TRUE,
pairwise.comparisons = TRUE,
ggstatsplot.layer = FALSE,
messages = FALSE
)
NU_MPRAGE_snrd.RepeatedMs.plot
# combining the individual plots into a single plot
ggstatsplot::combine_plots(
NU_MPRAGE_cnr.RepeatedMs.plot, NU_MPRAGE_fwhm.RepeatedMs.plot, NU_MPRAGE_snr.RepeatedMs.plot, NU_MPRAGE_snrd.RepeatedMs.plot,
nrow = 2,
title.text = "Northeastern Repeated Measures: \nMPRAGE IQMs Non-parametric Test",
caption.text = "Method: Non-parametric test from `ggstatsplot` R package",
title.size = 12,
caption.size = 10
)
# Mann-Whitney U test (nonparametric test)
PITT_MPRAGE_cnr.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_PITT,
x = Session,
y = MPRAGE_cnr,
xlab = "Session",
ylab = "CNR",
type = "np",
conf.level = 0.99,
ggstatsplot.layer = FALSE,
messages = FALSE
)
PITT_MPRAGE_cnr.RepeatedMs.plot
PITT_MPRAGE_fwhm.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_PITT,
x = Session,
y = MPRAGE_fwhm_avg,
xlab = "Session",
ylab = "FWHM",
type = "np",
conf.level = 0.99,
ggstatsplot.layer = FALSE,
messages = FALSE
)
PITT_MPRAGE_fwhm.RepeatedMs.plot
PITT_MPRAGE_snr.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_PITT,
x = Session,
y = MPRAGE_snr_total,
xlab = "Session",
ylab = "SNR",
type = "np",
conf.level = 0.99,
ggstatsplot.layer = FALSE,
messages = FALSE
)
PITT_MPRAGE_snr.RepeatedMs.plot
PITT_MPRAGE_snrd.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_PITT,
x = Session,
y = MPRAGE_snrd_total,
xlab = "Session",
ylab = "SNRd",
type = "np",
pairwise.annotation = "p-value",
outlier.tagging = TRUE,
outlier.label = SubID,
pairwise.comparisons = TRUE,
ggstatsplot.layer = FALSE,
messages = FALSE
)
PITT_MPRAGE_snrd.RepeatedMs.plot
# combining the individual plots into a single plot
ggstatsplot::combine_plots(
PITT_MPRAGE_cnr.RepeatedMs.plot, PITT_MPRAGE_fwhm.RepeatedMs.plot, PITT_MPRAGE_snr.RepeatedMs.plot, PITT_MPRAGE_snrd.RepeatedMs.plot,
nrow = 2,
title.text = "Pittsburgh Site Repeated Measures: \nMPRAGE IQMs Non-parametric Test",
caption.text = "Method: Non-parametric test from `ggstatsplot` R package",
title.size = 12,
caption.size = 10
)
DATA_Kansas_Skyra## SubID Session SITE Scanner MPRAGE_cnr MPRAGE_fwhm_avg
## 121 20014 MR1 KU KU Skyra 1.776900 4.036300
## 122 20014 MR2 KU KU Skyra 1.882200 4.037100
## 123 20018 MR2 KU KU Skyra 1.546000 4.074000
## 124 20018 MR1 KU KU Skyra 1.675100 4.037700
## 125 20021 MR2 KU KU Skyra 1.648900 4.318200
## 126 20021 MR1 KU KU Skyra 1.597900 4.115700
## 127 20022 MR1 KU KU Skyra 1.589100 3.945300
## 128 20022 MR2 KU KU Skyra 1.630000 3.985200
## 129 20025 MR2 KU KU Skyra 1.564300 4.227000
## 130 20025 MR1 KU KU Skyra 1.512700 4.188100
## 131 20027 MR2 KU KU Skyra 1.581100 4.114200
## 132 20027 MR1 KU KU Skyra 1.653200 4.052400
## 133 20028 MR2 KU KU Skyra 1.414900 3.916700
## 134 20028 MR1 KU KU Skyra 1.385200 3.965200
## 135 20033 MR2 KU KU Skyra 1.740900 4.215900
## 136 20033 MR1 KU KU Skyra 1.573600 4.255400
## 137 20034 MR2 KU KU Skyra 1.574100 4.091400
## 138 20034 MR1 KU KU Skyra 1.369600 4.130400
## 139 20035 MR1 KU KU Skyra 1.568000 4.053100
## 140 20035 MR2 KU KU Skyra 1.450300 4.122000
## 141 20041 MR2 KU KU Skyra 1.650700 3.885200
## 142 20041 MR1 KU KU Skyra 1.756200 3.808800
## 143 20045 MR1 KU KU Skyra 1.649000 4.176000
## 144 20045 MR2 KU KU Skyra 1.610200 4.045300
## 145 20050 MR2 KU KU Skyra 1.831900 3.975600
## 146 20050 MR1 KU KU Skyra 1.792500 3.859300
## 147 20057 MR2 KU KU Skyra 1.888600 3.915100
## 148 20057 MR1 KU KU Skyra 1.910700 3.928300
## 149 20058 MR1 KU KU Skyra 1.599900 4.166000
## 150 20058 MR2 KU KU Skyra 1.604500 4.051100
## 151 20062 MR1 KU KU Skyra 1.736300 4.047600
## 152 20062 MR2 KU KU Skyra 1.644400 4.051100
## 153 20065 MR1 KU KU Skyra 1.551700 4.005300
## 154 20065 MR2 KU KU Skyra 1.141700 3.866500
## 155 20070 MR2 KU KU Skyra 1.350500 4.080100
## 156 20070 MR1 KU KU Skyra 1.523600 3.947800
## 157 20075 MR1 KU KU Skyra 1.622500 4.049200
## 158 20075 MR2 KU KU Skyra 1.779100 4.155000
## 159 20077 MR2 KU KU Skyra 1.630200 4.130700
## 160 20077 MR1 KU KU Skyra 1.523500 4.009700
## 161 20080 MR2 KU KU Skyra 1.349800 4.037500
## 162 20080 MR1 KU KU Skyra 1.393700 3.872700
## 163 20082 MR1 KU KU Skyra 1.952600 4.109500
## 164 20082 MR2 KU KU Skyra 1.890600 4.008400
## 165 20084 MR1 KU KU Skyra 1.422200 3.973200
## 166 20084 MR2 KU KU Skyra 1.457800 4.193800
## 167 20085 MR1 KU KU Skyra 1.822800 4.091400
## 168 20085 MR2 KU KU Skyra 1.668500 4.068300
## 169 20087 MR1 KU KU Skyra 1.509700 4.399700
## 170 20087 MR2 KU KU Skyra 1.610800 4.252800
## 171 20088 MR2 KU KU Skyra 1.482800 4.210900
## 172 20088 MR1 KU KU Skyra 1.469500 4.167500
## 173 20089 MR2 KU KU Skyra 1.485500 3.877600
## 174 20089 MR1 KU KU Skyra 1.379700 3.769800
## 175 20091 MR1 KU KU Skyra 1.619900 3.956900
## 176 20091 MR2 KU KU Skyra 1.554600 3.957500
## 177 20093 MR2 KU KU Skyra 1.525400 4.123900
## 178 20093 MR1 KU KU Skyra 1.646300 4.279700
## 179 20096 MR2 KU KU Skyra 1.518600 3.833900
## 180 20096 MR1 KU KU Skyra 1.548200 3.882000
## 181 20099 MR2 KU KU Skyra 1.583200 3.973800
## 182 20099 MR1 KU KU Skyra 1.537400 3.959300
## 183 20101 MR2 KU KU Skyra 1.367000 3.965500
## 184 20101 MR1 KU KU Skyra 1.321100 4.078600
## 185 20104 MR2 KU KU Skyra 1.868300 3.873700
## 186 20104 MR1 KU KU Skyra 1.748100 3.794500
## 187 20108 MR1 KU KU Skyra 1.676100 3.862700
## 188 20108 MR2 KU KU Skyra 1.601800 3.867300
## 189 20110 MR1 KU KU Skyra 1.547300 3.929500
## 190 20110 MR2 KU KU Skyra 1.488500 4.040800
## 191 20116 MR1 KU KU Skyra 1.597800 4.311100
## 192 20116 MR2 KU KU Skyra 1.539300 4.417000
## 193 20118 MR1 KU KU Skyra 1.576900 3.916400
## 194 20118 MR2 KU KU Skyra 1.467400 4.020800
## 195 20120 MR2 KU KU Skyra 1.890900 4.197900
## 196 20120 MR1 KU KU Skyra 1.861600 4.177600
## 197 20121 MR1 KU KU Skyra 1.685500 4.202100
## 198 20121 MR2 KU KU Skyra 1.829400 4.245100
## 199 20122 MR1 KU KU Skyra 1.658600 3.889800
## 200 20122 MR2 KU KU Skyra 1.533900 3.956800
## 201 20123 MR2 KU KU Skyra 1.610600 3.986500
## 202 20123 MR1 KU KU Skyra 1.638900 3.967700
## 203 20132 MR2 KU KU Skyra 1.756800 3.953700
## 204 20132 MR1 KU KU Skyra 1.655300 3.978400
## 205 20133 MR1 KU KU Skyra 1.821000 4.183700
## 206 20133 MR2 KU KU Skyra 1.962100 4.209300
## 207 20138 MR2 KU KU Skyra 1.476000 4.185400
## 208 20138 MR1 KU KU Skyra 1.668800 4.203400
## 209 20140 MR2 KU KU Skyra 1.941500 4.008500
## 210 20140 MR1 KU KU Skyra 1.793100 3.948800
## 211 20142 MR1 KU KU Skyra 1.872300 4.110700
## 212 20142 MR2 KU KU Skyra 1.670500 4.326000
## 213 20145 MR2 KU KU Skyra 1.491000 4.064100
## 214 20145 MR1 KU KU Skyra 1.337900 4.099400
## 215 20146 MR2 KU KU Skyra 1.773800 4.099100
## 216 20146 MR1 KU KU Skyra 1.807200 3.950300
## 217 20161 MR2 KU KU Skyra 1.738900 3.999500
## 218 20161 MR1 KU KU Skyra 1.826900 3.921700
## 219 20162 MR2 KU KU Skyra 1.548100 3.990600
## 220 20162 MR1 KU KU Skyra 1.668400 3.884300
## 221 20165 MR1 KU KU Skyra 1.504800 3.963800
## 222 20165 MR2 KU KU Skyra 1.497100 3.904800
## 223 20171 MR1 KU KU Skyra 1.478100 3.889600
## 224 20171 MR2 KU KU Skyra 1.474600 3.999400
## 225 20175 MR2 KU KU Skyra 1.677502 4.324955
## 226 20175 MR1 KU KU Skyra 1.817300 4.249700
## 227 20181 MR2 KU KU Skyra 1.577757 3.913434
## 228 20181 MR1 KU KU Skyra 1.456500 4.152500
## 229 20188 MR1 KU KU Skyra 1.468700 4.004200
## 230 20188 MR2 KU KU Skyra 1.341200 4.076400
## 231 20217 MR1 KU KU Skyra 1.277700 3.840100
## 232 20217 MR2 KU KU Skyra 1.321800 3.826200
## MPRAGE_snr_total MPRAGE_snrd_total N
## 121 10.65610 55.61740 1
## 122 11.08000 58.66040 1
## 123 11.10110 71.39930 1
## 124 11.34120 70.32050 1
## 125 10.66120 68.32890 1
## 126 10.59500 64.19540 1
## 127 10.92100 74.40690 1
## 128 10.80890 65.38940 1
## 129 10.53490 76.16950 1
## 130 10.61310 70.99170 1
## 131 10.77090 70.71890 1
## 132 11.01440 66.20960 1
## 133 10.62540 65.28600 1
## 134 10.60400 64.53050 1
## 135 9.72420 54.83180 1
## 136 9.74920 68.96910 1
## 137 10.91690 64.69370 1
## 138 10.52030 70.04690 1
## 139 10.06050 61.21590 1
## 140 9.92910 60.68200 1
## 141 10.75690 65.35840 1
## 142 11.04540 57.42990 1
## 143 11.00790 67.73960 1
## 144 10.87940 67.76470 1
## 145 11.30240 71.15800 1
## 146 10.85770 64.15920 1
## 147 11.07200 59.32700 1
## 148 11.26220 57.80620 1
## 149 10.38980 64.01470 1
## 150 10.32330 55.83700 1
## 151 10.84140 76.55860 1
## 152 10.51120 64.64450 1
## 153 10.28400 70.27000 1
## 154 9.39760 28.79250 1
## 155 9.83250 82.18640 1
## 156 10.15100 92.00340 1
## 157 10.44750 71.22090 1
## 158 10.58270 81.74980 1
## 159 10.67870 50.87220 1
## 160 10.50550 54.47550 1
## 161 10.29380 62.60690 1
## 162 10.49540 64.64990 1
## 163 11.26310 52.84280 1
## 164 11.01320 53.03600 1
## 165 9.93820 79.51520 1
## 166 9.57370 74.36350 1
## 167 10.68650 64.39860 1
## 168 10.23500 78.62310 1
## 169 9.85940 52.03350 1
## 170 9.94800 51.47350 1
## 171 9.56200 55.90160 1
## 172 10.09150 63.93620 1
## 173 10.15430 74.69850 1
## 174 9.48910 72.18060 1
## 175 9.48460 64.50950 1
## 176 9.33000 55.74950 1
## 177 10.66130 84.46520 1
## 178 11.29370 77.92720 1
## 179 9.77770 61.80480 1
## 180 9.71590 59.89520 1
## 181 10.44120 70.43370 1
## 182 10.16890 71.24250 1
## 183 9.98210 60.45130 1
## 184 9.87370 62.65850 1
## 185 10.70490 59.61080 1
## 186 10.23250 65.05050 1
## 187 10.09970 74.39450 1
## 188 10.12060 79.78960 1
## 189 10.01030 75.40460 1
## 190 9.79500 74.95000 1
## 191 11.14300 80.71330 1
## 192 10.58020 69.82350 1
## 193 10.32900 63.60830 1
## 194 10.01970 60.92130 1
## 195 10.28640 66.56420 1
## 196 10.50940 62.87640 1
## 197 10.56220 64.59950 1
## 198 10.95030 65.86060 1
## 199 10.65070 72.84200 1
## 200 10.60200 71.72380 1
## 201 10.76110 58.07410 1
## 202 10.80990 56.28820 1
## 203 10.41890 66.86840 1
## 204 9.72690 39.99110 1
## 205 10.30810 62.71030 1
## 206 10.74570 54.34460 1
## 207 10.71490 69.74890 1
## 208 11.21150 64.39170 1
## 209 10.92590 60.05250 1
## 210 9.97660 47.37780 1
## 211 11.16010 68.29940 1
## 212 10.69660 61.98890 1
## 213 11.15040 64.22470 1
## 214 10.76990 69.65030 1
## 215 9.55040 46.99350 1
## 216 9.60330 48.25200 1
## 217 10.21220 94.68070 1
## 218 10.50070 90.46430 1
## 219 9.70710 52.52760 1
## 220 9.99110 50.01460 1
## 221 10.91740 66.69300 1
## 222 10.51490 65.40190 1
## 223 10.69900 64.53990 1
## 224 10.77850 66.33390 1
## 225 10.56387 56.16384 1
## 226 11.08420 59.75650 1
## 227 10.00178 66.27772 1
## 228 9.92920 62.67400 1
## 229 10.90420 58.49940 1
## 230 10.41320 59.49420 1
## 231 9.57040 70.16240 1
## 232 9.94020 80.21860 1# Mann-Whitney U test (nonparametric test)
KU_MPRAGE_cnr.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_Kansas_Skyra,
x = Session,
y = MPRAGE_cnr,
xlab = "Session",
ylab = "CNR",
type = "np",
conf.level = 0.99,
ggstatsplot.layer = FALSE,
messages = FALSE
)
KU_MPRAGE_cnr.RepeatedMs.plot
KU_MPRAGE_fwhm.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_Kansas_Skyra,
x = Session,
y = MPRAGE_fwhm_avg,
xlab = "Session",
ylab = "FWHM",
type = "np",
conf.level = 0.99,
ggstatsplot.layer = FALSE,
messages = FALSE
)
KU_MPRAGE_fwhm.RepeatedMs.plot
KU_MPRAGE_snr.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_Kansas_Skyra,
x = Session,
y = MPRAGE_snr_total,
xlab = "Session",
ylab = "SNR",
type = "np",
conf.level = 0.99,
ggstatsplot.layer = FALSE,
messages = FALSE
)
KU_MPRAGE_snr.RepeatedMs.plot
KU_MPRAGE_snrd.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_Kansas_Skyra,
x = Session,
y = MPRAGE_snrd_total,
xlab = "Session",
ylab = "SNRd",
type = "np",
pairwise.annotation = "p-value",
outlier.tagging = TRUE,
outlier.label = SubID,
pairwise.comparisons = TRUE,
ggstatsplot.layer = FALSE,
messages = FALSE
)
KU_MPRAGE_snrd.RepeatedMs.plot
# combining the individual plots into a single plot
ggstatsplot::combine_plots(
KU_MPRAGE_cnr.RepeatedMs.plot, KU_MPRAGE_fwhm.RepeatedMs.plot, KU_MPRAGE_snr.RepeatedMs.plot, KU_MPRAGE_snrd.RepeatedMs.plot,
nrow = 2,
title.text = "Kansas Site Repeated Measures: \nMPRAGE IQMs Non-parametric Test",
caption.text = "Method: Non-parametric test from `ggstatsplot` R package",
title.size = 12,
caption.size = 10
)
## Linear mixed-effects model fit by REML
## Data: Completed.Repeated.df
## Log-restricted-likelihood: 74.01776
## Fixed: MPRAGE_cnr ~ Scanner * Session
## (Intercept) ScannerNEU Prisma
## 1.61491250 0.21201417
## ScannerPITT Prisma 1 ScannerPITT Prisma 2
## 0.07153496 -0.14321250
## SessionMR2 ScannerNEU Prisma:SessionMR2
## -0.01245075 0.01920408
## ScannerPITT Prisma 1:SessionMR2 ScannerPITT Prisma 2:SessionMR2
## 0.04148784 0.04665075
##
## Random effects:
## Formula: ~1 | SubID
## (Intercept) Residual
## StdDev: 0.1621287 0.1231769
##
## Number of Observations: 292
## Number of Groups: 146## numDF denDF F-value p-value
## (Intercept) 1 142 12320.900 <.0001
## Scanner 3 142 9.923 <.0001
## Session 1 142 0.354 0.5527
## Scanner:Session 3 142 0.552 0.6479# Mann-Whitney U test (nonparametric test)
MPRAGE_cnr.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = Completed.Repeated.df,
x = Session,
y = MPRAGE_cnr,
xlab = "Session",
ylab = "MPRAGE_cnr",
type = "np",
conf.level = 0.99,
title = "Non-parametric Test",
package = "ggsci",
palette = "uniform_startrek",
ggtheme = ggthemes::theme_map(),
ggstatsplot.layer = FALSE,
messages = FALSE
)
MPRAGE_cnr.RepeatedMs.plot
MPRAGE_fwhm_avg.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = Completed.Repeated.df,
x = Session,
y = MPRAGE_fwhm_avg,
xlab = "Session",
ylab = "MPRAGE_fwhm_avg",
type = "np",
conf.level = 0.99,
title = "Non-parametric Test",
package = "ggsci",
palette = "uniform_startrek",
ggtheme = ggthemes::theme_map(),
ggstatsplot.layer = FALSE,
messages = FALSE
)
MPRAGE_fwhm_avg.RepeatedMs.plot
MPRAGE_snr_total.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = Completed.Repeated.df,
x = Session,
y = MPRAGE_snr_total,
xlab = "Session",
ylab = "MPRAGE_snr_total",
type = "np",
conf.level = 0.99,
title = "Non-parametric Test",
package = "ggsci",
palette = "uniform_startrek",
ggtheme = ggthemes::theme_map(),
ggstatsplot.layer = FALSE,
messages = FALSE
)
MPRAGE_snr_total.RepeatedMs.plot
MPRAGE_snrd_total.RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = Completed.Repeated.df,
x = Session,
y = MPRAGE_snrd_total,
xlab = "Session",
ylab = "MPRAGE_snrd_total",
type = "np",
conf.level = 0.99,
title = "Non-parametric Test",
package = "ggsci",
palette = "uniform_startrek",
ggtheme = ggthemes::theme_map(),
ggstatsplot.layer = FALSE,
messages = FALSE
)
MPRAGE_snrd_total.RepeatedMs.plot
Completed.Repeated.df$N<-1
BySession<-Completed.Repeated.df %>%
group_by(Scanner, Session) %>%
summarise( ScannerCount=sum(N),
meanCNR=mean(MPRAGE_cnr), sdCNR=sd(MPRAGE_cnr),
meanFWHM=mean(MPRAGE_fwhm_avg), sdFWHM=sd(MPRAGE_fwhm_avg),
meanSNR=mean(MPRAGE_snr_total), sdSNR=sd(MPRAGE_snr_total),
meanSNRd=mean(MPRAGE_snrd_total), sdSNRd=sd(MPRAGE_snrd_total))
BySession## # A tibble: 8 x 11
## # Groups: Scanner [4]
## Scanner Session ScannerCount meanCNR sdCNR meanFWHM sdFWHM meanSNR
## <fct> <fct> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 KU Sky… MR1 56 1.61 0.159 4.03 0.141 10.5
## 2 KU Sky… MR2 56 1.60 0.174 4.06 0.137 10.4
## 3 NEU Pr… MR1 30 1.83 0.250 4.12 0.136 10.6
## 4 NEU Pr… MR2 30 1.83 0.277 4.16 0.153 10.4
## 5 PITT P… MR1 59 1.69 0.174 4.14 0.144 10.1
## 6 PITT P… MR2 59 1.72 0.224 4.16 0.154 10.2
## 7 PITT P… MR1 1 1.47 NA 4.19 NA 9.41
## 8 PITT P… MR2 1 1.51 NA 4.26 NA 9.93
## # … with 3 more variables: sdSNR <dbl>, meanSNRd <dbl>, sdSNRd <dbl>## Step 3
DATA_NEU_Prisma<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="NEU Prisma",]
DATA_Kansas_Skyra<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="KU Skyra",]
DATA_PITT_Prisma_1<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="PITT Prisma 1",]
DATA_PITT_Prisma_2<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="PITT Prisma 2",]
NU_MPRAGEsnr_RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_NEU_Prisma,
x = Session,
y = MPRAGE_snr_total,
xlab = "Session",
ylab = "MPRAGE_snr_total",
type = "np",
conf.level = 0.99,
title = "Northeastern Repeated Measures Non-Parametric Test",
package = "ggsci",
palette = "uniform_startrek",
ggtheme = ggthemes::theme_map(),
ggstatsplot.layer = FALSE,
messages = FALSE
)
NU_MPRAGEsnr_RepeatedMs.plot
KU_MPRAGEsnr_RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_Kansas_Skyra,
x = Session,
y = MPRAGE_snr_total,
xlab = "Session",
ylab = "MPRAGE_snr_total",
type = "np",
conf.level = 0.99,
title = "Kansas Repeated Measures Non-Parametric Test",
package = "ggsci",
palette = "uniform_startrek",
ggtheme = ggthemes::theme_map(),
ggstatsplot.layer = FALSE,
messages = FALSE
)
KU_MPRAGEsnr_RepeatedMs.plot
Pitt1_MPRAGEsnr_RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_PITT_Prisma_1,
x = Session,
y = MPRAGE_snr_total,
xlab = "Session",
ylab = "MPRAGE_snr_total",
type = "np",
conf.level = 0.99,
title = "Pitt Prisma 1 Repeated Measures Non-Parametric Test",
package = "ggsci",
palette = "uniform_startrek",
ggtheme = ggthemes::theme_map(),
ggstatsplot.layer = FALSE,
messages = FALSE
)
Pitt1_MPRAGEsnr_RepeatedMs.plot
Pitt2_MPRAGEsnr_RepeatedMs.plot<- ggstatsplot::ggwithinstats(
data = DATA_PITT_Prisma_2,
x = Session,
y = MPRAGE_snr_total,
xlab = "Session",
ylab = "MPRAGE_snr_total",
type = "np",
conf.level = 0.99,
title = "Pitt Prisma 2 Repeated Measures Non-Parametric Test",
package = "ggsci",
palette = "uniform_startrek",
ggtheme = ggthemes::theme_map(),
ggstatsplot.layer = FALSE,
messages = FALSE
)## Warning: cannot compute confidence interval when all observations are zero
## or tied## Warning in stats::qt(p = 1 - (0.05/2), df = n - 1): NaNs produced
## Warning in stats::qt(p = 1 - (0.05/2), df = n - 1): NaNs producedPitt2_MPRAGEsnr_RepeatedMs.plot## Warning in max(data$density): no non-missing arguments to max; returning -
## Inf
Two-way repeated measures ANOVA
Assumptions:
- No missing values
- Normality of distribution (use the Shapiro-Wilk),
- Sphericity (e.g. equality of variance for all possible pairs of groups) in the experimental design.
- Groups contain no outliers
## Warning in as_grob.default(plot): Cannot convert object of class list into
## a grob.
Remove variance from scanner (Scale IQMs by Scanner) and complete repeated measures anova within subjects
## Step 3
DATA_NEU_Prisma_Z<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="NEU Prisma",]
DATA_Kansas_Skyra_Z<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="KU Skyra",]
DATA_PITT_Prisma_1_Z<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="PITT Prisma 1",]
DATA_PITT_Prisma_2_Z<-Completed.Repeated.df[Completed.Repeated.df$Scanner=="PITT Prisma 2",]
DATA_NEU_Prisma_Z[,5:8]<-scale(DATA_NEU_Prisma_Z[,5:8])
DATA_Kansas_Skyra_Z[,5:8]<-scale(DATA_Kansas_Skyra_Z[,5:8])
DATA_PITT_Prisma_1_Z[,5:8]<-scale(DATA_PITT_Prisma_1_Z[,5:8])
DATA_PITT_Prisma_2_Z[,5:8]<-scale(DATA_PITT_Prisma_2_Z[,5:8])
## Put it all back together now!
Complete_SiteZ.df<-rbind.data.frame(DATA_NEU_Prisma_Z,DATA_Kansas_Skyra_Z, DATA_PITT_Prisma_1_Z,DATA_PITT_Prisma_2_Z)
Complete_SiteZ.df<-Complete_SiteZ.df[!is.na(Complete_SiteZ.df$SubID),]
Complete_SiteZ.df$Session<-as.factor(Complete_SiteZ.df$Session)
rm(DATA_Kansas_Skyra_Z, DATA_NEU_Prisma_Z, DATA_PITT_Prisma_1_Z, DATA_PITT_Prisma_2_Z)
results.lme.scanner_sess_ind <- lme(MPRAGE_cnr ~ Scanner + Session, random=~1|SubID, data=Completed.Repeated.df)
anova(results.lme.scanner_sess_ind)## numDF denDF F-value p-value
## (Intercept) 1 145 12320.900 <.0001
## Scanner 3 142 9.923 <.0001
## Session 1 145 0.358 0.5508