Who am I (Dementia) Symptoms: PCA Analysis

Dementia

Dementia is an overall term describing a group of memory-deteriorating symptoms that associate one person’s ability of his or her everyday activities. Alzheimer’s disease is only accounts for 60% to 80 % of Dementia cases. Common dementia types:

Alzheimer

Alzheimer’s disease is a type of dementia causing problems with memory-deteriorating, thinking-deteriorating and behavior-deteriorating symptoms over time. Aging is necessary condition for Dementia but not sufficient condition for Alzheimer’s disease. Alzheimer’s disease does not necessarily occur for the old age. For example, approximately 200,000 Americans under the age of 65 have early-onset Alzheimer’s disease. The sixth leading cause of death in the United States is Alzheimer’s disease. In general, Alzheimer’s patient can survive from 4 to 20 years. All depend on age and other health conditions. The life-span of Alzheimer’s patient is 8 years on average. Currently, there is no cure for Alzheimer’s disease but some treatments for Alzheimer’s symptoms are available for stopping Alzheimer’s from progressing, temporarily decelerating the worsening of dementia symptoms and improving quality of life for Alzheimer’s patients and their caregivers. Early diagnostics and classification prediction are the current-best-solution approaches for Dementia.

Longitudinal MRI Dataset Description

Longitudinal MRI data: 150 subjects aged 60 to 96. All subjects are right-handed and scanned at least once.
‘Nondemented’ Group: 72 subjects
‘Demented’ Group: 64 subjects
Initial visit as ‘Nondemented’ but characterized as ‘Demented’ at later visit: 14 subjects

Variables	Descriptions
EDUC	Years of Education
SES	Socioeconomic Status
MMSE	Mini Mental State Examination
CDR	Clinical Dementia Rating
eTIV	Estimated Total Intracranial Volume
nWBV	Normalize Whole Brain Volume
ASF	Atlas Scaling Factor

Mini–Mental State Examination (MMSE)

The Mini–Mental State Examination (MMSE) is a 30-point questionnaire to measure cognitive impairment for screening-out dementia. The score estimation is
- to reflect the severity and progression of cognitive impairment,
- to follow-up the course of cognitive changes for each individual over time, and
- to provide a diagnosis for any particular type of Dementia disease.
The Mini–Mental State Examination (MMSE) is mainly to measure the degree of severity of Cognitive impairment for patients, ranging from mild to severe. such as
- the trouble degree of remembering,
- the trouble degree of learning new things,
- the trouble degree of concentration, and
- the trouble degree of decision-making in their everyday life.
The score estimation interpretation is list out in below Table

Methods	Score	Interpretation
SingleCutoff	< 24	Abnormal
RangeLessThan	< 21	Increased Odds of Dementia
RangeLessThan	< 25	Decreased Odds of Dementia
Education	= 21	Abnormal for 8th Grade Education
Education	< 23	Abnormal for High School Education
Education	< 24	Abnormal for College Education
Severity	24 - 30	No Cognitive Impairment
Severity	18 - 23	Mild Cognitive Impairment
Severity	0 - 17	Severe Cognitive Impairmen

Clinical Dementia Rating (CDR)

The CDR is a 5-point scale characterized six domains of cognitive and functional performance for Alzheimer disease and related dementia symptom. The six domains entail
- Memory,
- Orientation,
- Judgment & Problem Solving,
- Community Affairs,
- Home & Hobbies, and
- Personal Care.

Scores	Descriptions
0	Normal
0.5	Very Mild Dementia
1	Mild Dementia
2	Moderate Dementia
3	Severe Dementia

The below clinical dementia rating scale table is sourced from Morris (1993). This table provides descriptive entities to guide the clinician in making appropriate ratings based on interview data and clinical judgment.

The overall CDR score is calculated through an implicit algorithm. The final score is useful to characterize and track a patient’s level of impairment and dementia.

Estimated Total Intracranial Volume(eTIV)

Total intracranial volume (TIV/ICV) is a significant covariate indicator of various types of brain diseases by studying the volumetric analyses of brain regions. Good references lists belows:

Voevodskaya et al. (2014)
Malone et al. (2015)
Sargolzaei et al. (2015)
Klasson et al. (2018)

Atlas Scaling Factor (ASF)

Atlas Scaling Factor (ASF) is an atlas-based head size normalization technique to measure the standardized total intracranial volume for comparison, classification and predication. Good references lists below:

Fulton et al. (2019)
Mehmood et al. (2021)

Plot Gallery(longitudinal)

oasis_longitudinal.csv

Summary

df_longit = na.omit(as.data.frame(data_longit))
df_longit$Group <- as.factor(df_longit$Group)
df_longit$Visit <- as.factor(df_longit$Visit)
df_longit$SES <- as.factor(df_longit$SES)
df_longit$MF <- as.factor(df_longit$MF)
df_longit$Hand <- as.factor(df_longit$Hand)
str(df_longit)

'data.frame':   354 obs. of  15 variables:
 $ Subject ID: chr  "OAS2_0001" "OAS2_0001" "OAS2_0004" "OAS2_0004" ...
 $ MRI ID    : chr  "OAS2_0001_MR1" "OAS2_0001_MR2" "OAS2_0004_MR1" "OAS2_0004_MR2" ...
 $ Group     : Factor w/ 3 levels "Converted","Demented",..: 3 3 3 3 3 3 3 3 3 2 ...
 $ Visit     : Factor w/ 5 levels "1","2","3","4",..: 1 2 1 2 1 2 3 1 2 1 ...
 $ MR Delay  : num  0 457 0 538 0 ...
 $ MF        : Factor w/ 2 levels "F","M": 2 2 1 1 2 2 2 1 1 2 ...
 $ Hand      : Factor w/ 1 level "R": 1 1 1 1 1 1 1 1 1 1 ...
 $ Age       : num  87 88 88 90 80 83 85 93 95 68 ...
 $ EDUC      : num  14 14 18 18 12 12 12 14 14 12 ...
 $ SES       : Factor w/ 5 levels "1","2","3","4",..: 2 2 3 3 4 4 4 2 2 2 ...
 $ MMSE      : num  27 30 28 27 28 29 30 30 29 27 ...
 $ CDR       : num  0 0 0 0 0 0.5 0 0 0 0.5 ...
 $ eTIV      : num  1987 2004 1215 1200 1689 ...
 $ nWBV      : num  0.696 0.681 0.71 0.718 0.712 0.711 0.705 0.698 0.703 0.806 ...
 $ ASF       : num  0.883 0.876 1.444 1.462 1.039 ...
 - attr(*, "spec")=
  .. cols(
  ..   `Subject ID` = col_character(),
  ..   `MRI ID` = col_character(),
  ..   Group = col_character(),
  ..   Visit = col_double(),
  ..   `MR Delay` = col_double(),
  ..   MF = col_character(),
  ..   Hand = col_character(),
  ..   Age = col_double(),
  ..   EDUC = col_double(),
  ..   SES = col_double(),
  ..   MMSE = col_double(),
  ..   CDR = col_double(),
  ..   eTIV = col_double(),
  ..   nWBV = col_double(),
  ..   ASF = col_double()
  .. )
 - attr(*, "na.action")= 'omit' Named int [1:19] 3 4 5 11 12 13 135 136 208 209 ...
  ..- attr(*, "names")= chr [1:19] "3" "4" "5" "11" ...

summary(df_longit)

  Subject ID           MRI ID                  Group     Visit  
 Length:354         Length:354         Converted  : 37   1:142  
 Class :character   Class :character   Demented   :127   2:137  
 Mode  :character   Mode  :character   Nondemented:190   3: 55  
                                                         4: 14  
                                                         5:  6  
                                                                
    MR Delay      MF      Hand         Age             EDUC       SES    
 Min.   :   0.0   F:204   R:354   Min.   :60.00   Min.   : 6.00   1: 88  
 1st Qu.:   0.0   M:150           1st Qu.:71.00   1st Qu.:12.00   2:103  
 Median : 559.5                   Median :77.00   Median :15.00   3: 82  
 Mean   : 601.4                   Mean   :77.03   Mean   :14.70   4: 74  
 3rd Qu.: 882.5                   3rd Qu.:82.00   3rd Qu.:16.75   5:  7  
 Max.   :2639.0                   Max.   :98.00   Max.   :23.00          
      MMSE            CDR              eTIV           nWBV       
 Min.   : 4.00   Min.   :0.0000   Min.   :1106   Min.   :0.6440  
 1st Qu.:27.00   1st Qu.:0.0000   1st Qu.:1358   1st Qu.:0.6990  
 Median :29.00   Median :0.0000   Median :1470   Median :0.7290  
 Mean   :27.41   Mean   :0.2712   Mean   :1490   Mean   :0.7299  
 3rd Qu.:30.00   3rd Qu.:0.5000   3rd Qu.:1595   3rd Qu.:0.7570  
 Max.   :30.00   Max.   :2.0000   Max.   :2004   Max.   :0.8370  
      ASF       
 Min.   :0.876  
 1st Qu.:1.100  
 Median :1.194  
 Mean   :1.194  
 3rd Qu.:1.292  
 Max.   :1.587

df_modified = df_longit[, c(-1, -2)]

Group

Group_Age

longit_Group_Age <- ggplot(df_longit, aes(x = Group, y = Age, fill = Group)) + geom_violin() +
    geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_Group_Age

Group_Edu

longit_Group_Edu <- ggplot(df_longit, aes(x = Group, y = EDUC, fill = Group)) + geom_violin() +
    geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_Group_Edu

Group_MMSE

longit_Group_MMSE <- ggplot(df_longit, aes(x = Group, y = MMSE, fill = Group)) +
    geom_violin() + geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_Group_MMSE

Group_CDR

longit_Group_CDR <- ggplot(df_longit, aes(x = Group, y = CDR, fill = Group)) + geom_violin() +
    geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_Group_CDR

Group_eTIV

longit_Group_eTIV <- ggplot(df_longit, aes(x = Group, y = eTIV, fill = Group)) +
    geom_violin() + geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_Group_eTIV

Group_nWBV

longit_Group_nWBV <- ggplot(df_longit, aes(x = Group, y = nWBV, fill = Group)) +
    geom_violin() + geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_Group_nWBV

Group_ASF

longit_Group_ASF <- ggplot(df_longit, aes(x = Group, y = ASF, fill = Group)) + geom_violin() +
    geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_Group_ASF

MF

MF_SES

longit_MF_SES <- ggplot(df_longit, aes(x = MF, y = SES, fill = MF)) + geom_violin() +
    geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_MF_SES

MF_MMSE

longit_MF_MMSE <- ggplot(df_longit, aes(x = MF, y = MMSE, fill = MF)) + geom_violin() +
    geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_MF_MMSE

MF_CDR

longit_MF_CDR <- ggplot(df_longit, aes(x = MF, y = CDR, fill = MF)) + geom_violin() +
    geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_MF_CDR

MF_eTIV

longit_MF_eTIV <- ggplot(df_longit, aes(x = MF, y = eTIV, fill = MF)) + geom_violin() +
    geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_MF_eTIV

MF_nWBV

longit_MF_nWBV <- ggplot(df_longit, aes(x = MF, y = nWBV, fill = MF)) + geom_violin() +
    geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_MF_nWBV

MF_ASF

longit_MF_ASF <- ggplot(df_longit, aes(x = MF, y = ASF, fill = MF)) + geom_violin() +
    geom_boxplot(width = 0.1) + theme(legend.position = "none")
longit_MF_ASF

ScatterPlot

data = df_longit[, c("CDR", "eTIV", "nWBV", "ASF")]
corr <- abs(cor(data))  # Correlation in absolute value
corr

            CDR       eTIV      nWBV        ASF
CDR  1.00000000 0.05236148 0.3500857 0.06341298
eTIV 0.05236148 1.00000000 0.2066682 0.98902995
nWBV 0.35008571 0.20666821 1.0000000 0.21114974
ASF  0.06341298 0.98902995 0.2111497 1.00000000

colors <- dmat.color(corr)
order <- order.single(corr)

cpairs(data, order, panel.colors = colors, gap = 0.5, main = "Sorted and colored variables by correlation")

pairs(~Group + MF + SES + EDUC + Age + CDR + eTIV + nWBV + ASF, data = df_longit,
    col = "blue")

PCA

data4 = df_longit[, c("CDR", "eTIV", "nWBV", "ASF")]
cor.mat <- round(cor(data4), 4)

corrplot(cor.mat, type = "upper", order = "hclust", tl.col = "black", tl.srt = 45)

chart.Correlation(data4, histogram = TRUE, pch = 19)

res.pca <- PCA(data4, graph = F)
eigenvalues <- res.pca$eig
eigenvalues

       eigenvalue percentage of variance cumulative percentage of variance
comp 1 2.09971778             52.4929445                          52.49294
comp 2 1.25757433             31.4393582                          83.93230
comp 3 0.63180008             15.7950021                          99.72730
comp 4 0.01090781              0.2726952                         100.00000

res.pca$var$coord

          Dim.1      Dim.2       Dim.3         Dim.4
CDR  -0.2418549  0.8110732  0.53260285  8.067998e-04
eTIV -0.9595790 -0.2638679  0.06428662  7.381819e-02
nWBV  0.4419949 -0.6822595  0.58237653 -6.176814e-05
ASF   0.9618070  0.2542248 -0.06956363  7.387846e-02

res.pca$var$cos2

          Dim.1      Dim.2       Dim.3        Dim.4
CDR  0.05849379 0.65783977 0.283665791 6.509259e-07
eTIV 0.92079183 0.06962627 0.004132770 5.449126e-03
nWBV 0.19535953 0.46547804 0.339162425 3.815303e-09
ASF  0.92507262 0.06463025 0.004839099 5.458027e-03

res.pca$var$contrib

         Dim.1     Dim.2      Dim.3        Dim.4
CDR   2.785793 52.310210 44.8980299 5.967523e-03
eTIV 43.853123  5.536553  0.6541262 4.995620e+01
nWBV  9.304085 37.013958 53.6819214 3.497772e-05
ASF  44.056998  5.139279  0.7659225 5.003780e+01

barplot(eigenvalues[, 2], names.arg = 1:nrow(eigenvalues), main = "Variances", xlab = "Principal Components",
    ylab = "Percentage of variances", col = "green")
# Add connected line segments to the plot
lines(x = 1:nrow(eigenvalues), eigenvalues[, 2], type = "b", pch = 19, col = "red")

fviz_pca_var(res.pca, col.var = "contrib") + scale_color_gradient2(low = "green",
    mid = "blue", high = "red", midpoint = 25) + theme_bw()

fviz_pca_biplot(res.pca, label = "var", habillage = df_longit$Group, addEllipses = TRUE,
    ellipse.level = 0.95, ggtheme = theme_minimal())

fviz_pca_biplot(res.pca, label = "var", habillage = df_longit$MF, addEllipses = TRUE,
    ellipse.level = 0.95, ggtheme = theme_minimal())

fviz_pca_biplot(res.pca, label = "var", habillage = df_longit$SES, addEllipses = TRUE,
    ellipse.level = 0.95, ggtheme = theme_minimal(), title = "PCA Biplot: Groups of Socioeconomic Status (SES)")

References

Fulton, Lawrence V., Diane Dolezel, Jordan Harrop, Yan Yan, and Christopher P. Fulton. 2019. “Classification of Alzheimer’s Disease with and Without Imagery Using Gradient Boosted Machines and ResNet-50.” Brain Sciences 9 (9). https://doi.org/10.3390/brainsci9090212.

Klasson, Niklas, Erik Olsson, Carl Eckerström, Helge Malmgren, and Anders Wallin. 2018. “Estimated Intracranial Volume from FreeSurfer Is Biased by Total Brain Volume.” European Radiology Experimental 2 (PMC6143491): 24. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6143491/.

Malone, Ian B., Kelvin K. Leung, Shona Clegg, Josephine Barnes, Jennifer L. Whitwell, John Ashburner, Nick C. Fox, and Gerard R. Ridgway. 2015. “Accurate Automatic Estimation of Total Intracranial Volume: A Nuisance Variable with Less Nuisance.” NeuroImage 104: 366–72. https://doi.org/https://doi.org/10.1016/j.neuroimage.2014.09.034.

Mehmood, Atif, Shuyuan Yang, Zhixi Feng, Min Wang, A. L. Smadi Ahmad, Rizwan Khan, Muazzam Maqsood, and Muhammad Yaqub. 2021. “A Transfer Learning Approach for Early Diagnosis of Alzheimer’s Disease on MRI Images.” Neuroscience 460: 43–52. https://www.sciencedirect.com/science/article/pii/S0306452221000075.

Morris, John C. 1993. “The Clinical Dementia Rating (CDR).” Neurology 43 (11): 2412-2412-a. https://doi.org/10.1212/WNL.43.11.2412-a.

Sargolzaei, Saman, Arman Sargolzaei, Mercedes Cabrerizo, Gang Chen, Mohammed Goryawala, Shirin Noei, Qi Zhou, Ranjan Duara, Warren Barker, and Malek Adjouadi. 2015. “A Practical Guideline for Intracranial Volume Estimation in Patients with Alzheimer’s Disease.” BMC Bioinformatics 16 Suppl 7 (25953026): S8–8. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4423585/.

Voevodskaya, Olga, Andrew Simmons, Richard Nordenskjöld, Joel Kullberg, Håkan Ahlström, Lars Lind, Lars-Olof Wahlund, et al. 2014. “The Effects of Intracranial Volume Adjustment Approaches on Multiple Regional MRI Volumes in Healthy Aging and Alzheimer’s Disease.” Frontiers in Aging Neuroscience 6: 264.