R/MotilityFunctions.R
In sebastiz/PhysiMineR: Functions to extract, clean, manipulate and analyse upper gastrointestinal physiological parameters
Defines functions
MotilityTimeSeries
MyImpedanceDataWithHRM
HRMDiagnoses
SymptomsExtraction
HRMCleanUp1
HRMAndDiag
HRMAndSwallows
dataHRM
Documented in
dataHRM
HRMAndDiag
HRMAndSwallows
HRMCleanUp1
HRMDiagnoses
MotilityTimeSeries
MyImpedanceDataWithHRM
SymptomsExtraction
#Motility functions
#### Data acquisition ####
#This depends on forming a connection to the PhysiPop database. In windows the path is this:
#channel <- odbcConnectAccess("S:\\Gastroenterology\\Seb\\JavaPortableLauncher\\PhysiPopDONOTTOUCH\\Physiology6.mdb")
#' dataHRM
#' This acquires HRM data from the database
#' @param channel odbcConnectAccess connection defined at the start of this file from RODBC function (for windows)
#' @importFrom RODBC sqlQuery
#' @keywords HRM data acquisition
#' @export
#' @examples #dataHRM(channel)
dataHRM<-function(channel){
data <- sqlQuery(channel , "SELECT PatientData.*, HRMImportMain.*
FROM PatientData INNER JOIN HRMImportMain ON PatientData.HospNum_Id = HRMImportMain.HospNum_Id")
return(data)
}
#' HRMAndSwallows
#' This acquires HRMSwallow data from the upper GI database
#' @param channel odbcConnectAccess connection defined at the start of this file from RODBC function (for windows)
#' @keywords HRMAndSwallows acquisition
#' @export
#' @examples #HRMAndSwallows(channel)
HRMAndSwallows<-function(channel){
data <- sqlQuery( channel , "SELECT PatientData.*, HRMImportSwallows.*, HRMImportMain.*
FROM PatientData INNER JOIN (HRMImportMain INNER JOIN HRMImportSwallows ON HRMImportMain.HRM_Id = HRMImportSwallows.HRM_Id) ON PatientData.HospNum_Id = HRMImportMain.HospNum_Id
WHERE HRMImportMain.HRM_Id=HRMImportSwallows.HRM_Id")
return(data)
}
#' HRMAndDiag
#' This acquires HRM and associated diagnosis data from the upper GI database
#' @param channel odbcConnectAccess connection defined at the start of this file from RODBC function (for windows)
#' @keywords HRM and Diag
#' @export
#' @examples #HRMAndDiag(channel)
HRMAndDiag<-function(channel){
data <- sqlQuery( channel , "SELECT DISTINCT HRMImportMain.*, Diag.IndicANDHx, Diag.*, PatientData.*
FROM (PatientData INNER JOIN Diag ON PatientData.HospNum_Id = Diag.HospNum_Id) INNER JOIN HRMImportMain ON PatientData.HospNum_Id = HRMImportMain.HospNum_Id
WHERE HRMImportMain.VisitDate=Diag.VisitDate")
return(data)
}
########################### Clean the motility data #################################################################################
#' HRMCleanUp1
#' This cleans HRM data by making sure all the data is in the correct format
#' @param x dataframe
#' @keywords HRM CleanUp
#' @importFrom lubridate parse_date_time
#' @export
#' @examples #HRMCleanUp1(x)
HRMCleanUp1<-function(x){
try(x$Gender<-gsub("(Male|Female).*","\\1",x$Gender))
try(x$Gender<-gsub("null:","",x$Gender))
#Format the columns so they are correct
#First need to convert the tibble to a dataframe
x<-data.frame(x,stringsAsFactors = FALSE)
#x<-lapply(x, function(y) { if(is.character(y)) gsub("NA:", "", x) else x })
#x<-lapply(x, function(y) { if(is.character(y)) gsub(":NA", "", x) else x })
#x<-lapply(x, function(y) { if(is.character(y)) gsub("N_A", "", x) else x })
#x<-lapply(x, function(y) { if(is.character(y)) gsub("NA", "", x) else x })
#Get rid of funny values from merged rows by gsub
x<-lapply(x,function(y) gsub("NA:","",y))
x<-lapply(x,function(y) gsub("N_A","",y))
x<-lapply(x,function(y) gsub(":NA","",y))
x<-lapply(x,function(y) gsub("NA",NA,y))
i1 <- grepl("Num\\d+", names(x))
x[i1] <- lapply(x[i1], as.numeric)
#Need to get rid of NA: in merged rows and convert numeric and digit to numbers
i1 <- grepl("cm|mmHg", names(x))
x[i1] <- lapply(x[i1], as.numeric)
#Do the same for digital numbers
i1 <- grepl("\\d+\\.\\d+", names(x))
x[i1] <- lapply(x[i1], as.numeric)
i1 <- grepl("^\\d*$", x)
x[i1] <- lapply(x[i1], as.numeric)
#Do the same for numbers only to avoid hospital numbers
i1 <- grepl("^\\d{1,2}$", names(x))
x[i1] <- lapply(x[i1], as.numeric)
#Just convert to the most appropriate
x<-lapply(x, type.convert, as.is = TRUE)
#Clean the demographics
is.date <- function(x) inherits(x, 'Date')
if(!is.date(x$VisitDate)){
x$VisitDate<-as.character(x$VisitDate)
x$VisitDate<-gsub("(^\\d{2}_\\d{2}_)(\\d{2}$)","\\120\\2",x$VisitDate)
x$VisitDate<-as.Date(x$VisitDate,"%d_%m_%Y")
}
#Get rid of anything that is after the first date
x$DOBAge<-gsub("(.*?\\d{4}).*","\\1",x$DOBAge)
#Standardise teh separators between numbers:
x$DOBAge<-gsub("-|\\.","_",x$DOBAge)
#Sort out the years which have been abbreviated to only two digits
x$DOBAge<-gsub("(^\\d{2}_\\d{2}_)(\\d{2}($|\\s))","\\119\\2",x$DOBAge)
#Get rid of extra characters after the date
x$DOBAge<-gsub("(^\\d{2}_\\d{2}_\\d{4}).*","\\1",x$DOBAge)
#Get rid of extra characters before the date
x$DOBAge<-gsub(".*(\\d{2}_\\d{2}_\\d{4})","\\1",x$DOBAge)
x$DOBAge<-parse_date_time(x = x$DOBAge,orders = c("d b Y", "m d Y","d m Y"))
#Convert dates with named months
x$DOBAge<-as.character(x$DOBAge)
x$DOBAge<-as.Date(x$DOBAge,"%Y-%m-%d")
x$Age<-x$VisitDate-x$DOBAge
x$Age<-difftime(x$VisitDate,x$DOBAge,units="days")/366.25
x$Age<-as.numeric(x$Age)
x$Hiatalhernia<-gsub("^.*:","",x$Hiatalhernia)
#This lot doesnt seem to want to be converted so do it manually:
x$Distallatency<-as.numeric(x$Distallatency)
x$failedChicagoClassification<-as.numeric(x$failedChicagoClassification)
x$panesophagealpressurization<-as.numeric(x$panesophagealpressurization)
x$largebreaks<- as.numeric(x$largebreaks)
x$Simultaneous<-as.numeric(x$Simultaneous)
x$prematurecontraction<-as.numeric(x$prematurecontraction)
x$rapidcontraction<-as.numeric(x$rapidcontraction)
x$smallbreaks<- as.numeric(x$smallbreaks)
x$Hiatalhernia<-as.character((x$Hiatalhernia))
#Convert other columns to numeric
x$Distallatency<-as.numeric(as.character((x$Distallatency)))
x$LOS_relax<-ifelse(((x$ResidualmeanmmHg-x$BasalrespiratoryminmmHg/x$BasalrespiratorymeanmmHg)*100)<90,"NonRelaxLOS","NormalRelaxLOS")
x$LOS_relax<-as.factor(x$LOS_relax)
x$LowerOesoph<-ifelse(x$BasalrespiratoryminmmHg<4.7&x$Hiatalhernia=="Yes","HypotensiveLOSWithHH",
ifelse(x$BasalrespiratoryminmmHg<4.7,"HypotensiveLOS",
ifelse(x$Hiatalhernia=="Yes","HHOnly","Normal")))
x$LowerOesoph<-as.factor(x$LowerOesoph)
x<-data.frame(x,stringsAsFactors = FALSE)
return(x)
}
########################### HRM Symptom Subset Prepare #################################################################################
#'SymptomsExtraction
#'
#' SymptomsExtraction
#' This extracts the symptoms from a pre-specified column into its own column per symptom. Usually these are extracted from the
#' associated diagnosis table as the HRM main table doesnt keep free text stuff
#' @param x dataframe
#' @param y the column of interest
#' @keywords Symptom
#' @export
#' @examples #SymptomsExtraction(x,IndicANDHx)
SymptomsExtraction<-function(x,y){
x$Dysphagia<-ifelse(grepl(".*[Dd]ysph.*",x[,y],perl=TRUE)|grepl(".*[Oo]dyn.*",x[,y],perl=TRUE)|grepl(".*[Ss]tuck.*",x[,y],perl=TRUE)|grepl(".*[Ss]tick.*",x[,y],perl=TRUE),"Dysphagia","No")
x$Heartburn<-ifelse(grepl(".*[Hh]eart.*",x[,y],perl=TRUE)|grepl(".*[Rr]eflu.*",x[,y],perl=TRUE)|grepl(".*[Rr]etro.*",x[,y],perl=TRUE)|grepl(".*[Bb]urn.*",x[,y],perl=TRUE),"Heartburn","No")
x$Throat<-ifelse(grepl(".*[Tt]hroat.*",x[,y],perl=TRUE)|grepl(".*[Nn]eck.*",x[,y],perl=TRUE),"Throat","No")
x$Cough<-ifelse(grepl(".*[Cc]ough.*",x[,y],perl=TRUE)|grepl(".*[Cc]hok.*",x[,y],perl=TRUE),"Cough","No")
x$ChestPain<-ifelse(grepl(".*[Ch]est.*",x[,y],perl=TRUE),"ChestPain","No")
x$AbdoPain<-ifelse(grepl(".*[Ss]tomach.*",x[,y],perl=TRUE)|grepl(".*[Ee]piga.*",x[,y],perl=TRUE)|grepl(".*[Aa]bdom.*",x[,y],perl=TRUE),"AbdoPain","No")
x$Hoarseness<-ifelse(grepl(".*[Hh]oarse.*",x[,y],perl=TRUE),"Hoarseness","No")
x$Regurgitation<-ifelse(grepl(".*[Rr]egur.*",x[,y],perl=TRUE)|grepl(".*[Tt]aste.*",x[,y],perl=TRUE),"Regurgitation","No")
x$Vomiting<-ifelse(grepl(".*[Vv]omit.*",x[,y],perl=TRUE),"Vomiting","No")
x$Belch<-ifelse(grepl(".*[Bb]elch.*",x[,y],perl=TRUE)|grepl(".*[Bb]urp.*",x[,y],perl=TRUE),"Belch","No")
x$AllSymptoms<-paste(x$Dysphagia,",",x$Heartburn,",",x$Throat,",",x$Cough,",",x$ChestPain,",",x$AbdoPain,",",x$Hoarseness,",",x$Regurgitation,",",x$Vomiting,",",x$Belch,",")
x$AllSymptoms<-gsub("NO,","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub(",NO","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub("NO","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub("NO","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub(" ","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub(",{2,}",",",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub("^,","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub(",$","",x$AllSymptoms,ignore.case=T)
x$Dysphagia<-NULL
x$Heartburn<-NULL
x$Throat<-NULL
x$Cough<-NULL
x$ChestPain<-NULL
x$AbdoPain<-NULL
x$Hoarseness<-NULL
x$Regurgitation<-NULL
x$Vomiting<-NULL
x$Belch<-NULL
return (x)
}
########################### Categorise the HRM diagnoses #################################################################################
#' HRMDiagnoses
#' This creates diagnoses from the HRM raw data based on the Chicago classification v4 (need to check)
#' @param x dataframe
#' @keywords HRM diagnoses
#' @export
#' @examples #HRMDiagnoses(x)
HRMDiagnoses<-function(x){
x$dx<-ifelse(x$ResidualmeanmmHg>15&x$failedChicagoClassification==100&!is.na(x$ResidualmeanmmHg)&!is.na(x$failedChicagoClassification),"AchalasiaType1",
ifelse(x$ResidualmeanmmHg>=15&!is.na(x$ResidualmeanmmHg)&x$prematurecontraction>=20,"AchalasiaType2",
ifelse(x$ResidualmeanmmHg>=15&!is.na(x$ResidualmeanmmHg)&x$panesophagealpressurization>=20,"AchalasiaType3",
ifelse(x$ResidualmeanmmHg>=15&!is.na(x$ResidualmeanmmHg)&x$failedChicagoClassification>0,"EGOO",
ifelse(x$ResidualmeanmmHg<15&x$ResidualmeanmmHg>10&!is.na(x$ResidualmeanmmHg)&x$failedChicagoClassification==100&!is.na(x$failedChicagoClassification),"PossibleAchalasia",
ifelse(x$ResidualmeanmmHg>=15&!is.na(x$ResidualmeanmmHg),"AchalasiaType2or3orEGOO",
ifelse(x$ResidualmeanmmHg<=15&x$failedChicagoClassification==100&!is.na(x$ResidualmeanmmHg)&!is.na(x$failedChicagoClassification),"AbsentPeristalsis",
ifelse(x$ResidualmeanmmHg<=15&(x$prematurecontraction>=20|x$Simultaneous>=20|x$Distallatency<4.5)&x$DistalcontractileintegralmeanmmHgcms>=450&!is.na(x$ResidualmeanmmHg)&(!is.na(x$prematurecontraction)&!is.na(x$Simultaneous))&!is.na(x$DistalcontractileintegralmeanmmHgcms),"DES",
ifelse(x$ResidualmeanmmHg<=15&(x$DistalcontractileintegralmeanmmHgcms>=8000)|(x$DistalcontractileintegralmeanmmHgcms>=8000|x$DistalcontractileintegralmeanmmHgcms>=8000)&!is.na(x$ResidualmeanmmHg)&!is.na(x$DistalcontractileintegralmeanmmHgcms),"JackHammer",
ifelse(x$ResidualmeanmmHg<=15& ((x$failedChicagoClassification<100&x$failedChicagoClassification>50)|grepl("IOM|neffective oesoph",x$WholeReport))&!is.na(x$ResidualmeanmmHg)&!is.na(x$DistalcontractileintegralmeanmmHgcms),"IOM",
#ifelse(x$ResidualmeanmmHg<15&x$Contractilefrontvelocitycms>9&x$Distallatency>=4.5&!is.na(x$ResidualmeanmmHg)&!is.na(x$Contractilefrontvelocitycms)&!is.na(x$Distallatency),"RapidContraction",
#ifelse(x$ResidualmeanmmHg<15&(x$DistalcontractileintegralmeanmmHgcms>=5000|x$DistalcontractileintegralmeanmmHgcms>=5000)&x$Distallatency>=4.5&!is.na(x$ResidualmeanmmHg)&!is.na(x$Distallatency)&!is.na(x$DistalcontractileintegralmeanmmHgcms),"HypertensivePeristalsis",
#ifelse(x$ResidualmeanmmHg<=15&x$smallbreaks>=30&x$largebreaks>=20&!is.na(x$ResidualmeanmmHg)&!is.na(x$smallbreaks)&!is.na(x$largebreaks),"WeakPeristalsis",
ifelse(x$ResidualmeanmmHg<15&x$largebreaks>=50&!is.na(x$ResidualmeanmmHg)&!is.na(x$failedChicagoClassification),"FrequentFailedPeristalsis","Normal")))))))))))
return(x)
}
#Note Ineffective esophageal motility is diagnosed when >50% of swallows is ineffective, that is either failed (DistalcontractileintegralmeanmmHgcms <100mmHg cm s) or weak (DistalcontractileintegralmeanmmHgcms 100–450 mmHg cm s).
#Fragmented peristalsis is defined as >50% of swallows with a large break (>5 cm) and not matching criteria for ineffective esophageal motility
##### Merging with another table ####
#' Merge HRM and Impedance
#' This merges the overall impedance dataset with the HRM dataset
#' @param x the whole impedance dataset
#' @param y the whole HRM dataset
#' @keywords HRMAndSwallows acquisition
#' @export
#' @examples
#' #dataHRM<-dataHRM(channel)
#' #dataImp2<-dataImp2(channel)
#' #dataImp_Symp<-dataImp_Symp(channel)
#'
#' #dataImpWhole<-dataImpWhole(dataImp2,dataImp_Symp)
#' #MyImpedanceDataWithHRM(dataImpWhole,dataHRM)
MyImpedanceDataWithHRM<-function(x,y){
MyImpedanceDataWithHRM<-merge(x,y,by=c("HospNum_Id"),all=TRUE)
return(MyImpedanceDataWithHRM)
}
########################### Some metrics #################################################################################
#' MotilityTimeSeries
#' Plots the Tests over time. Probably redundant
#' @param x dataframe
#' @keywords HRM Motility
#' @export
#' @examples #MotilityTimeSeries(x)
#'
MotilityTimeSeries <- function(x) {
xTimePlot<-x %>%
mutate(month=format(VisitDate,"%m"), year= format(VisitDate,"%Y")) %>%
group_by(month,year)
xTimePlot$MergeCol<-paste("01",xTimePlot$month, xTimePlot$year, sep=" ")
xTimePlot<-data.frame(table(xTimePlot$MergeCol))
names(xTimePlot)<-c("Date","Freq")
xTimePlot<-subset(xTimePlot,!xTimePlot$Date=="NA/NA")
xTimePlot$Date<-as.Date(xTimePlot$Date,format="%d %m %Y")
PlotName<-deparse(substitute(x))
print(PlotName)
xTimePlot<-xTimePlot[order(xTimePlot$Date),]
myplot<-ggplot(xTimePlot) +
geom_point(aes(Date, Freq, color = "red"))+
labs(title=PlotName) +
scale_color_manual("",labels = c("SVStart", "SVEnd"), values = c("blue", "red")) +
xlab("Date") +
ylab("Frequency") +
theme(axis.text.x=element_text(angle=-90)) +
theme(legend.position="top")
}
#----------------------------Sandbox
#Define multiple logistic regression and exploratory analysis first
#see https://rcompanion.org/rcompanion/e_07.html
# library(PerformanceAnalytics)
#chart.Correlation(HRMAll_Diag4[13:19], method="spearman",histogram=TRUE,pch=16)
##### Symptom vs Parameter
#eg. Question:What proportion of NERD is related to acid vs non acid. ie What proportion of {condition} is related to {parameter}
#1. Chronic Cough Is Associated With Long Breaks in Esophageal Peristaltic Integrity on High-resolution Manometry. J Neurogastroenterol Motil. 2018 Jul 30;24(3):387-394. doi: 10.5056/jnm17126.
#################{symptom} is associated with {parameter}
#2. Factors predictive of gastroesophageal reflux disease and esophageal motility disorders in patients with non-cardiac chest pain. Gomez Cifuentes J1, Lopez R2, Thota PN3. Scand J Gastroenterol. 2018 Jun;53(6):643-649. doi: 10.1080/00365521.2018.1452975.
###########{symptom subgroup} {parameter} predicts {condition}
##### Parameter vs Condition
#1. Swallow-induced esophageal shortening in patients without hiatal hernia is associated with gastroesophageal reflux. #Masuda T, Singhal S, Akimoto S, Bremner RM, Mittal SK. Dis Esophagus. 2018 May 1;31(5). doi: 10.1093/dote/dox152. PMID: 29293978
###########{{parameter} is associated with {condition}
#2. Comparison of esophageal motility in gastroesophageal reflux disease with and without globus sensation. Tang Y, Huang J, Zhu Y, Qian A, Xu B, Yao W.#Rev Esp Enferm Dig. 2017 Dec;109(12):850-855. doi: 10.17235/reed.2017.4449/2016.
###########{{symptom subgroup} {parameter} is associated with {condition}
#3. High-resolution manometry in patients with and without globus pharyngeus and/or symptoms of laryngopharyngeal reflux. #Ding H, Duan Z, Yang D, Zhang Z, Wang L, Sun X, Yao Y, Lin X, Yang H, Wang S, Chen JDZ. BMC Gastroenterol. 2017 Oct 23;17(1):109. doi: 10.1186/s12876-017-0666-x. PMID: 29061118 Free PMC Article{symptom subgroup} {parameter} is associated with {condition}
###########{{symptom subgroup} {parameter} is associated with {condition}
#4. Distribution of Esophageal Motor Disorders in Diabetic Patients With Dysphagia. George NS, Rangan V, Geng Z, Khan F, Kichler A, Gabbard S, Ganocy S, Fass R. J Clin Gastroenterol. 2017 Nov/Dec;51(10):890-895. doi: 10.1097/MCG.0000000000000894. PMID: 28746079
###########{{symptom subgroup} {parameter} is associated with {condition}
#5 Upper esophageal sphincter (UES) metrics on high-resolution manometry (HRM) differentiate achalasia subtypes.
###########{{condition subgroup} {parameter} is associated with {condition}
###### Parameter pre and post an intervention
#1. Esophageal Motility after Extensive Circumferential Endoscopic Submucosal Dissection for Superficial Esophageal Cancer.#Digestion. 2018 Jun 5;98(3):153-160. doi: 10.1159/000487751. [Epub ahead of print]
#2. Impact of thoracic surgery on esophageal motor function-Evaluation by high resolution manometry.Wäsche A, Kandulski A, Malfertheiner P, Riedel S, Zardo P, Hachenberg T, Schreiber J.J Thorac Dis. 2017 Jun;9(6):1557-1564. doi: 10.21037/jtd.2017.05.43.PMID: 28740669 Free PMC Article
###### Condition vs Parameter- mainly descriptive
#1.GERD: Presence and Size of Hiatal Hernia Influence Clinical Presentation, Esophageal Function, Reflux Profile, and Degree of Mucosal Injury. # Schlottmann F, Andolfi C, Herbella FA, Rebecchi F, Allaix ME, Patti MG.Am Surg. 2018 Jun 1;84(6):978-982.
#2.Esophageal Motor Disorders Are a Strong and Independant Associated Factor of Barrett's Esophagus. #Bazin C, Benezech A, Alessandrini M, Grimaud JC, Vitton V. #J Neurogastroenterol Motil. 2018 Apr 30;24(2):216-225. doi: 10.5056/jnm17090. #PMID: 29605977 Free PMC Article
###########{{condition} is associated with {parameter}
#3. Esophagogastric junction outflow obstruction is often associated with coexistent abnormal esophageal body motility and abnormal bolus transit. #Zheng E, Gideon RM, Sloan J, Katz PO. Dis Esophagus. 2017 Oct 1;30(10):1-4. doi: 10.1093/dote/dox066.
###########{{parameter} is associated with {condition}
###########{Filtered subgroup
#Predictive variables
#Subclassifying algorithms
###########{function(subgroup,parameter OR symptom OR condition,parameter OR symptom OR condition ){
###########{}
# What {Parameter/condition/symptom} in this subgroup are associated with {Parameter/condition/symptom}?
#
# library(rattle)
#
#
#
# Filtered subgroup:Non cardiac_Chest_pain
#
# Is a associated with b
# eg Specific<-function(Non-subgroup,one Parameter, achalasia){
#
#
# }
#
# Is there a variable in a subgroup that can predict b
# Multifac<-function(Non-cardiac_Chest_pain,all_device_parameters, achalasia){
#
#
#
# }
#
Symptom<-as.list(c("Non-cardiac chest pain",
"Heartburn",
"Regurgitation",
"Hoarse voice",
"Dysphagia",
"Cough",
"Globus",
"Throat clearing",
"abdominal pain"))
Parameterz<-as.list(c("DistalLESfromnarescm",
"LESmidpointfromnarescm",
"ProximalLESfromnarescm",
"LESlengthcm",
"EsophageallengthLESUEScenterscm",
"PIPfromnarescm",
"IntraabdominalLESlengthcm",
"Hiatalhernia",
"BasalrespiratoryminmmHg",
"BasalrespiratorymeanmmHg",
"ResidualmeanmmHg",
"UESMeanResidLocationcenterfrnarescm",
"ResidMeanbasalpressuremmHg",
"ResidMeanresidualpressuremmHg",
"Numberofswallowsevaluated",
"DistalcontractileintegralhighestmmHgcms",
"DistalcontractileintegralmeanmmHgcms",
"Contractilefrontvelocitycms",
"IntraboluspressureATLESRmmHg",
"Distallatency",
"failedChicagoClassification",
"panesophagealpressurization",
"largebreaks",
"Simultaneous",
"prematurecontraction",
"rapidcontraction",
"smallbreaks",
"VisitDate",
"DOBAge"))
Condition<-as.list(c("Achalasia type 1",
"Achalasia type 2",
"Achalasia type 3",
"EGOO",
"DES",
"Jackhammer",
"Aperistaltic",
"Frequent failed Peristalsis",
"Ineffective oesophageal motility"))
paste("Investigating the association between ",Condition," and ",Parameterz, "in patients with ",Symptom )
paste("Investigating the association between ",Symptom," and ",Parameterz, "in patients with ",Condition )
paste("Investigating the association between ",Symptom," and ",Condition, "in patients with ", Parameterz)
#
# #Patients with condition and a symptom analysis of parameters
# #ie which parameters are significantly elevated in patients with specific symptoms and a specific condition
#
#
# ###### Condition description (ie subclassification)
#
# #1. Clinical and manometric characteristics of patients with oesophagogastric outflow obstruction: towards a new classification George Triadafilopoulos, John O Clarke
#
# ##### Device vs device
# #
# #
# # Both vs future outcomes
# # Both vs future conditions
sebastiz/PhysiMineR documentation built on Oct. 3, 2023, 3:46 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions MotilityTimeSeries MyImpedanceDataWithHRM HRMDiagnoses SymptomsExtraction HRMCleanUp1 HRMAndDiag HRMAndSwallows dataHRM
Documented in dataHRM HRMAndDiag HRMAndSwallows HRMCleanUp1 HRMDiagnoses MotilityTimeSeries MyImpedanceDataWithHRM SymptomsExtraction
#Motility functions
#### Data acquisition ####
#This depends on forming a connection to the PhysiPop database. In windows the path is this:
#channel <- odbcConnectAccess("S:\\Gastroenterology\\Seb\\JavaPortableLauncher\\PhysiPopDONOTTOUCH\\Physiology6.mdb")
#' dataHRM
#' This acquires HRM data from the database
#' @param channel odbcConnectAccess connection defined at the start of this file from RODBC function (for windows)
#' @importFrom RODBC sqlQuery
#' @keywords HRM data acquisition
#' @export
#' @examples #dataHRM(channel)
dataHRM<-function(channel){
data <- sqlQuery(channel , "SELECT PatientData.*, HRMImportMain.*
FROM PatientData INNER JOIN HRMImportMain ON PatientData.HospNum_Id = HRMImportMain.HospNum_Id")
return(data)
}
#' HRMAndSwallows
#' This acquires HRMSwallow data from the upper GI database
#' @param channel odbcConnectAccess connection defined at the start of this file from RODBC function (for windows)
#' @keywords HRMAndSwallows acquisition
#' @export
#' @examples #HRMAndSwallows(channel)
HRMAndSwallows<-function(channel){
data <- sqlQuery( channel , "SELECT PatientData.*, HRMImportSwallows.*, HRMImportMain.*
FROM PatientData INNER JOIN (HRMImportMain INNER JOIN HRMImportSwallows ON HRMImportMain.HRM_Id = HRMImportSwallows.HRM_Id) ON PatientData.HospNum_Id = HRMImportMain.HospNum_Id
WHERE HRMImportMain.HRM_Id=HRMImportSwallows.HRM_Id")
return(data)
}
#' HRMAndDiag
#' This acquires HRM and associated diagnosis data from the upper GI database
#' @param channel odbcConnectAccess connection defined at the start of this file from RODBC function (for windows)
#' @keywords HRM and Diag
#' @export
#' @examples #HRMAndDiag(channel)
HRMAndDiag<-function(channel){
data <- sqlQuery( channel , "SELECT DISTINCT HRMImportMain.*, Diag.IndicANDHx, Diag.*, PatientData.*
FROM (PatientData INNER JOIN Diag ON PatientData.HospNum_Id = Diag.HospNum_Id) INNER JOIN HRMImportMain ON PatientData.HospNum_Id = HRMImportMain.HospNum_Id
WHERE HRMImportMain.VisitDate=Diag.VisitDate")
return(data)
}
########################### Clean the motility data #################################################################################
#' HRMCleanUp1
#' This cleans HRM data by making sure all the data is in the correct format
#' @param x dataframe
#' @keywords HRM CleanUp
#' @importFrom lubridate parse_date_time
#' @export
#' @examples #HRMCleanUp1(x)
HRMCleanUp1<-function(x){
try(x$Gender<-gsub("(Male|Female).*","\\1",x$Gender))
try(x$Gender<-gsub("null:","",x$Gender))
#Format the columns so they are correct
#First need to convert the tibble to a dataframe
x<-data.frame(x,stringsAsFactors = FALSE)
#x<-lapply(x, function(y) { if(is.character(y)) gsub("NA:", "", x) else x })
#x<-lapply(x, function(y) { if(is.character(y)) gsub(":NA", "", x) else x })
#x<-lapply(x, function(y) { if(is.character(y)) gsub("N_A", "", x) else x })
#x<-lapply(x, function(y) { if(is.character(y)) gsub("NA", "", x) else x })
#Get rid of funny values from merged rows by gsub
x<-lapply(x,function(y) gsub("NA:","",y))
x<-lapply(x,function(y) gsub("N_A","",y))
x<-lapply(x,function(y) gsub(":NA","",y))
x<-lapply(x,function(y) gsub("NA",NA,y))
i1 <- grepl("Num\\d+", names(x))
x[i1] <- lapply(x[i1], as.numeric)
#Need to get rid of NA: in merged rows and convert numeric and digit to numbers
i1 <- grepl("cm|mmHg", names(x))
x[i1] <- lapply(x[i1], as.numeric)
#Do the same for digital numbers
i1 <- grepl("\\d+\\.\\d+", names(x))
x[i1] <- lapply(x[i1], as.numeric)
i1 <- grepl("^\\d*$", x)
x[i1] <- lapply(x[i1], as.numeric)
#Do the same for numbers only to avoid hospital numbers
i1 <- grepl("^\\d{1,2}$", names(x))
x[i1] <- lapply(x[i1], as.numeric)
#Just convert to the most appropriate
x<-lapply(x, type.convert, as.is = TRUE)
#Clean the demographics
is.date <- function(x) inherits(x, 'Date')
if(!is.date(x$VisitDate)){
x$VisitDate<-as.character(x$VisitDate)
x$VisitDate<-gsub("(^\\d{2}_\\d{2}_)(\\d{2}$)","\\120\\2",x$VisitDate)
x$VisitDate<-as.Date(x$VisitDate,"%d_%m_%Y")
}
#Get rid of anything that is after the first date
x$DOBAge<-gsub("(.*?\\d{4}).*","\\1",x$DOBAge)
#Standardise teh separators between numbers:
x$DOBAge<-gsub("-|\\.","_",x$DOBAge)
#Sort out the years which have been abbreviated to only two digits
x$DOBAge<-gsub("(^\\d{2}_\\d{2}_)(\\d{2}($|\\s))","\\119\\2",x$DOBAge)
#Get rid of extra characters after the date
x$DOBAge<-gsub("(^\\d{2}_\\d{2}_\\d{4}).*","\\1",x$DOBAge)
#Get rid of extra characters before the date
x$DOBAge<-gsub(".*(\\d{2}_\\d{2}_\\d{4})","\\1",x$DOBAge)
x$DOBAge<-parse_date_time(x = x$DOBAge,orders = c("d b Y", "m d Y","d m Y"))
#Convert dates with named months
x$DOBAge<-as.character(x$DOBAge)
x$DOBAge<-as.Date(x$DOBAge,"%Y-%m-%d")
x$Age<-x$VisitDate-x$DOBAge
x$Age<-difftime(x$VisitDate,x$DOBAge,units="days")/366.25
x$Age<-as.numeric(x$Age)
x$Hiatalhernia<-gsub("^.*:","",x$Hiatalhernia)
#This lot doesnt seem to want to be converted so do it manually:
x$Distallatency<-as.numeric(x$Distallatency)
x$failedChicagoClassification<-as.numeric(x$failedChicagoClassification)
x$panesophagealpressurization<-as.numeric(x$panesophagealpressurization)
x$largebreaks<- as.numeric(x$largebreaks)
x$Simultaneous<-as.numeric(x$Simultaneous)
x$prematurecontraction<-as.numeric(x$prematurecontraction)
x$rapidcontraction<-as.numeric(x$rapidcontraction)
x$smallbreaks<- as.numeric(x$smallbreaks)
x$Hiatalhernia<-as.character((x$Hiatalhernia))
#Convert other columns to numeric
x$Distallatency<-as.numeric(as.character((x$Distallatency)))
x$LOS_relax<-ifelse(((x$ResidualmeanmmHg-x$BasalrespiratoryminmmHg/x$BasalrespiratorymeanmmHg)*100)<90,"NonRelaxLOS","NormalRelaxLOS")
x$LOS_relax<-as.factor(x$LOS_relax)
x$LowerOesoph<-ifelse(x$BasalrespiratoryminmmHg<4.7&x$Hiatalhernia=="Yes","HypotensiveLOSWithHH",
ifelse(x$BasalrespiratoryminmmHg<4.7,"HypotensiveLOS",
ifelse(x$Hiatalhernia=="Yes","HHOnly","Normal")))
x$LowerOesoph<-as.factor(x$LowerOesoph)
x<-data.frame(x,stringsAsFactors = FALSE)
return(x)
}
########################### HRM Symptom Subset Prepare #################################################################################
#'SymptomsExtraction
#'
#' SymptomsExtraction
#' This extracts the symptoms from a pre-specified column into its own column per symptom. Usually these are extracted from the
#' associated diagnosis table as the HRM main table doesnt keep free text stuff
#' @param x dataframe
#' @param y the column of interest
#' @keywords Symptom
#' @export
#' @examples #SymptomsExtraction(x,IndicANDHx)
SymptomsExtraction<-function(x,y){
x$Dysphagia<-ifelse(grepl(".*[Dd]ysph.*",x[,y],perl=TRUE)|grepl(".*[Oo]dyn.*",x[,y],perl=TRUE)|grepl(".*[Ss]tuck.*",x[,y],perl=TRUE)|grepl(".*[Ss]tick.*",x[,y],perl=TRUE),"Dysphagia","No")
x$Heartburn<-ifelse(grepl(".*[Hh]eart.*",x[,y],perl=TRUE)|grepl(".*[Rr]eflu.*",x[,y],perl=TRUE)|grepl(".*[Rr]etro.*",x[,y],perl=TRUE)|grepl(".*[Bb]urn.*",x[,y],perl=TRUE),"Heartburn","No")
x$Throat<-ifelse(grepl(".*[Tt]hroat.*",x[,y],perl=TRUE)|grepl(".*[Nn]eck.*",x[,y],perl=TRUE),"Throat","No")
x$Cough<-ifelse(grepl(".*[Cc]ough.*",x[,y],perl=TRUE)|grepl(".*[Cc]hok.*",x[,y],perl=TRUE),"Cough","No")
x$ChestPain<-ifelse(grepl(".*[Ch]est.*",x[,y],perl=TRUE),"ChestPain","No")
x$AbdoPain<-ifelse(grepl(".*[Ss]tomach.*",x[,y],perl=TRUE)|grepl(".*[Ee]piga.*",x[,y],perl=TRUE)|grepl(".*[Aa]bdom.*",x[,y],perl=TRUE),"AbdoPain","No")
x$Hoarseness<-ifelse(grepl(".*[Hh]oarse.*",x[,y],perl=TRUE),"Hoarseness","No")
x$Regurgitation<-ifelse(grepl(".*[Rr]egur.*",x[,y],perl=TRUE)|grepl(".*[Tt]aste.*",x[,y],perl=TRUE),"Regurgitation","No")
x$Vomiting<-ifelse(grepl(".*[Vv]omit.*",x[,y],perl=TRUE),"Vomiting","No")
x$Belch<-ifelse(grepl(".*[Bb]elch.*",x[,y],perl=TRUE)|grepl(".*[Bb]urp.*",x[,y],perl=TRUE),"Belch","No")
x$AllSymptoms<-paste(x$Dysphagia,",",x$Heartburn,",",x$Throat,",",x$Cough,",",x$ChestPain,",",x$AbdoPain,",",x$Hoarseness,",",x$Regurgitation,",",x$Vomiting,",",x$Belch,",")
x$AllSymptoms<-gsub("NO,","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub(",NO","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub("NO","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub("NO","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub(" ","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub(",{2,}",",",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub("^,","",x$AllSymptoms,ignore.case=T)
x$AllSymptoms<-gsub(",$","",x$AllSymptoms,ignore.case=T)
x$Dysphagia<-NULL
x$Heartburn<-NULL
x$Throat<-NULL
x$Cough<-NULL
x$ChestPain<-NULL
x$AbdoPain<-NULL
x$Hoarseness<-NULL
x$Regurgitation<-NULL
x$Vomiting<-NULL
x$Belch<-NULL
return (x)
}
########################### Categorise the HRM diagnoses #################################################################################
#' HRMDiagnoses
#' This creates diagnoses from the HRM raw data based on the Chicago classification v4 (need to check)
#' @param x dataframe
#' @keywords HRM diagnoses
#' @export
#' @examples #HRMDiagnoses(x)
HRMDiagnoses<-function(x){
x$dx<-ifelse(x$ResidualmeanmmHg>15&x$failedChicagoClassification==100&!is.na(x$ResidualmeanmmHg)&!is.na(x$failedChicagoClassification),"AchalasiaType1",
ifelse(x$ResidualmeanmmHg>=15&!is.na(x$ResidualmeanmmHg)&x$prematurecontraction>=20,"AchalasiaType2",
ifelse(x$ResidualmeanmmHg>=15&!is.na(x$ResidualmeanmmHg)&x$panesophagealpressurization>=20,"AchalasiaType3",
ifelse(x$ResidualmeanmmHg>=15&!is.na(x$ResidualmeanmmHg)&x$failedChicagoClassification>0,"EGOO",
ifelse(x$ResidualmeanmmHg<15&x$ResidualmeanmmHg>10&!is.na(x$ResidualmeanmmHg)&x$failedChicagoClassification==100&!is.na(x$failedChicagoClassification),"PossibleAchalasia",
ifelse(x$ResidualmeanmmHg>=15&!is.na(x$ResidualmeanmmHg),"AchalasiaType2or3orEGOO",
ifelse(x$ResidualmeanmmHg<=15&x$failedChicagoClassification==100&!is.na(x$ResidualmeanmmHg)&!is.na(x$failedChicagoClassification),"AbsentPeristalsis",
ifelse(x$ResidualmeanmmHg<=15&(x$prematurecontraction>=20|x$Simultaneous>=20|x$Distallatency<4.5)&x$DistalcontractileintegralmeanmmHgcms>=450&!is.na(x$ResidualmeanmmHg)&(!is.na(x$prematurecontraction)&!is.na(x$Simultaneous))&!is.na(x$DistalcontractileintegralmeanmmHgcms),"DES",
ifelse(x$ResidualmeanmmHg<=15&(x$DistalcontractileintegralmeanmmHgcms>=8000)|(x$DistalcontractileintegralmeanmmHgcms>=8000|x$DistalcontractileintegralmeanmmHgcms>=8000)&!is.na(x$ResidualmeanmmHg)&!is.na(x$DistalcontractileintegralmeanmmHgcms),"JackHammer",
ifelse(x$ResidualmeanmmHg<=15& ((x$failedChicagoClassification<100&x$failedChicagoClassification>50)|grepl("IOM|neffective oesoph",x$WholeReport))&!is.na(x$ResidualmeanmmHg)&!is.na(x$DistalcontractileintegralmeanmmHgcms),"IOM",
#ifelse(x$ResidualmeanmmHg<15&x$Contractilefrontvelocitycms>9&x$Distallatency>=4.5&!is.na(x$ResidualmeanmmHg)&!is.na(x$Contractilefrontvelocitycms)&!is.na(x$Distallatency),"RapidContraction",
#ifelse(x$ResidualmeanmmHg<15&(x$DistalcontractileintegralmeanmmHgcms>=5000|x$DistalcontractileintegralmeanmmHgcms>=5000)&x$Distallatency>=4.5&!is.na(x$ResidualmeanmmHg)&!is.na(x$Distallatency)&!is.na(x$DistalcontractileintegralmeanmmHgcms),"HypertensivePeristalsis",
#ifelse(x$ResidualmeanmmHg<=15&x$smallbreaks>=30&x$largebreaks>=20&!is.na(x$ResidualmeanmmHg)&!is.na(x$smallbreaks)&!is.na(x$largebreaks),"WeakPeristalsis",
ifelse(x$ResidualmeanmmHg<15&x$largebreaks>=50&!is.na(x$ResidualmeanmmHg)&!is.na(x$failedChicagoClassification),"FrequentFailedPeristalsis","Normal")))))))))))
return(x)
}
#Note Ineffective esophageal motility is diagnosed when >50% of swallows is ineffective, that is either failed (DistalcontractileintegralmeanmmHgcms <100mmHg cm s) or weak (DistalcontractileintegralmeanmmHgcms 100–450 mmHg cm s).
#Fragmented peristalsis is defined as >50% of swallows with a large break (>5 cm) and not matching criteria for ineffective esophageal motility
##### Merging with another table ####
#' Merge HRM and Impedance
#' This merges the overall impedance dataset with the HRM dataset
#' @param x the whole impedance dataset
#' @param y the whole HRM dataset
#' @keywords HRMAndSwallows acquisition
#' @export
#' @examples
#' #dataHRM<-dataHRM(channel)
#' #dataImp2<-dataImp2(channel)
#' #dataImp_Symp<-dataImp_Symp(channel)
#'
#' #dataImpWhole<-dataImpWhole(dataImp2,dataImp_Symp)
#' #MyImpedanceDataWithHRM(dataImpWhole,dataHRM)
MyImpedanceDataWithHRM<-function(x,y){
MyImpedanceDataWithHRM<-merge(x,y,by=c("HospNum_Id"),all=TRUE)
return(MyImpedanceDataWithHRM)
}
########################### Some metrics #################################################################################
#' MotilityTimeSeries
#' Plots the Tests over time. Probably redundant
#' @param x dataframe
#' @keywords HRM Motility
#' @export
#' @examples #MotilityTimeSeries(x)
#'
MotilityTimeSeries <- function(x) {
xTimePlot<-x %>%
mutate(month=format(VisitDate,"%m"), year= format(VisitDate,"%Y")) %>%
group_by(month,year)
xTimePlot$MergeCol<-paste("01",xTimePlot$month, xTimePlot$year, sep=" ")
xTimePlot<-data.frame(table(xTimePlot$MergeCol))
names(xTimePlot)<-c("Date","Freq")
xTimePlot<-subset(xTimePlot,!xTimePlot$Date=="NA/NA")
xTimePlot$Date<-as.Date(xTimePlot$Date,format="%d %m %Y")
PlotName<-deparse(substitute(x))
print(PlotName)
xTimePlot<-xTimePlot[order(xTimePlot$Date),]
myplot<-ggplot(xTimePlot) +
geom_point(aes(Date, Freq, color = "red"))+
labs(title=PlotName) +
scale_color_manual("",labels = c("SVStart", "SVEnd"), values = c("blue", "red")) +
xlab("Date") +
ylab("Frequency") +
theme(axis.text.x=element_text(angle=-90)) +
theme(legend.position="top")
}
#----------------------------Sandbox
#Define multiple logistic regression and exploratory analysis first
#see https://rcompanion.org/rcompanion/e_07.html
# library(PerformanceAnalytics)
#chart.Correlation(HRMAll_Diag4[13:19], method="spearman",histogram=TRUE,pch=16)
##### Symptom vs Parameter
#eg. Question:What proportion of NERD is related to acid vs non acid. ie What proportion of {condition} is related to {parameter}
#1. Chronic Cough Is Associated With Long Breaks in Esophageal Peristaltic Integrity on High-resolution Manometry. J Neurogastroenterol Motil. 2018 Jul 30;24(3):387-394. doi: 10.5056/jnm17126.
#################{symptom} is associated with {parameter}
#2. Factors predictive of gastroesophageal reflux disease and esophageal motility disorders in patients with non-cardiac chest pain. Gomez Cifuentes J1, Lopez R2, Thota PN3. Scand J Gastroenterol. 2018 Jun;53(6):643-649. doi: 10.1080/00365521.2018.1452975.
###########{symptom subgroup} {parameter} predicts {condition}
##### Parameter vs Condition
#1. Swallow-induced esophageal shortening in patients without hiatal hernia is associated with gastroesophageal reflux. #Masuda T, Singhal S, Akimoto S, Bremner RM, Mittal SK. Dis Esophagus. 2018 May 1;31(5). doi: 10.1093/dote/dox152. PMID: 29293978
###########{{parameter} is associated with {condition}
#2. Comparison of esophageal motility in gastroesophageal reflux disease with and without globus sensation. Tang Y, Huang J, Zhu Y, Qian A, Xu B, Yao W.#Rev Esp Enferm Dig. 2017 Dec;109(12):850-855. doi: 10.17235/reed.2017.4449/2016.
###########{{symptom subgroup} {parameter} is associated with {condition}
#3. High-resolution manometry in patients with and without globus pharyngeus and/or symptoms of laryngopharyngeal reflux. #Ding H, Duan Z, Yang D, Zhang Z, Wang L, Sun X, Yao Y, Lin X, Yang H, Wang S, Chen JDZ. BMC Gastroenterol. 2017 Oct 23;17(1):109. doi: 10.1186/s12876-017-0666-x. PMID: 29061118 Free PMC Article{symptom subgroup} {parameter} is associated with {condition}
###########{{symptom subgroup} {parameter} is associated with {condition}
#4. Distribution of Esophageal Motor Disorders in Diabetic Patients With Dysphagia. George NS, Rangan V, Geng Z, Khan F, Kichler A, Gabbard S, Ganocy S, Fass R. J Clin Gastroenterol. 2017 Nov/Dec;51(10):890-895. doi: 10.1097/MCG.0000000000000894. PMID: 28746079
###########{{symptom subgroup} {parameter} is associated with {condition}
#5 Upper esophageal sphincter (UES) metrics on high-resolution manometry (HRM) differentiate achalasia subtypes.
###########{{condition subgroup} {parameter} is associated with {condition}
###### Parameter pre and post an intervention
#1. Esophageal Motility after Extensive Circumferential Endoscopic Submucosal Dissection for Superficial Esophageal Cancer.#Digestion. 2018 Jun 5;98(3):153-160. doi: 10.1159/000487751. [Epub ahead of print]
#2. Impact of thoracic surgery on esophageal motor function-Evaluation by high resolution manometry.Wäsche A, Kandulski A, Malfertheiner P, Riedel S, Zardo P, Hachenberg T, Schreiber J.J Thorac Dis. 2017 Jun;9(6):1557-1564. doi: 10.21037/jtd.2017.05.43.PMID: 28740669 Free PMC Article
###### Condition vs Parameter- mainly descriptive
#1.GERD: Presence and Size of Hiatal Hernia Influence Clinical Presentation, Esophageal Function, Reflux Profile, and Degree of Mucosal Injury. # Schlottmann F, Andolfi C, Herbella FA, Rebecchi F, Allaix ME, Patti MG.Am Surg. 2018 Jun 1;84(6):978-982.
#2.Esophageal Motor Disorders Are a Strong and Independant Associated Factor of Barrett's Esophagus. #Bazin C, Benezech A, Alessandrini M, Grimaud JC, Vitton V. #J Neurogastroenterol Motil. 2018 Apr 30;24(2):216-225. doi: 10.5056/jnm17090. #PMID: 29605977 Free PMC Article
###########{{condition} is associated with {parameter}
#3. Esophagogastric junction outflow obstruction is often associated with coexistent abnormal esophageal body motility and abnormal bolus transit. #Zheng E, Gideon RM, Sloan J, Katz PO. Dis Esophagus. 2017 Oct 1;30(10):1-4. doi: 10.1093/dote/dox066.
###########{{parameter} is associated with {condition}
###########{Filtered subgroup
#Predictive variables
#Subclassifying algorithms
###########{function(subgroup,parameter OR symptom OR condition,parameter OR symptom OR condition ){
###########{}
# What {Parameter/condition/symptom} in this subgroup are associated with {Parameter/condition/symptom}?
#
# library(rattle)
#
#
#
# Filtered subgroup:Non cardiac_Chest_pain
#
# Is a associated with b
# eg Specific<-function(Non-subgroup,one Parameter, achalasia){
#
#
# }
#
# Is there a variable in a subgroup that can predict b
# Multifac<-function(Non-cardiac_Chest_pain,all_device_parameters, achalasia){
#
#
#
# }
#
Symptom<-as.list(c("Non-cardiac chest pain",
"Heartburn",
"Regurgitation",
"Hoarse voice",
"Dysphagia",
"Cough",
"Globus",
"Throat clearing",
"abdominal pain"))
Parameterz<-as.list(c("DistalLESfromnarescm",
"LESmidpointfromnarescm",
"ProximalLESfromnarescm",
"LESlengthcm",
"EsophageallengthLESUEScenterscm",
"PIPfromnarescm",
"IntraabdominalLESlengthcm",
"Hiatalhernia",
"BasalrespiratoryminmmHg",
"BasalrespiratorymeanmmHg",
"ResidualmeanmmHg",
"UESMeanResidLocationcenterfrnarescm",
"ResidMeanbasalpressuremmHg",
"ResidMeanresidualpressuremmHg",
"Numberofswallowsevaluated",
"DistalcontractileintegralhighestmmHgcms",
"DistalcontractileintegralmeanmmHgcms",
"Contractilefrontvelocitycms",
"IntraboluspressureATLESRmmHg",
"Distallatency",
"failedChicagoClassification",
"panesophagealpressurization",
"largebreaks",
"Simultaneous",
"prematurecontraction",
"rapidcontraction",
"smallbreaks",
"VisitDate",
"DOBAge"))
Condition<-as.list(c("Achalasia type 1",
"Achalasia type 2",
"Achalasia type 3",
"EGOO",
"DES",
"Jackhammer",
"Aperistaltic",
"Frequent failed Peristalsis",
"Ineffective oesophageal motility"))
paste("Investigating the association between ",Condition," and ",Parameterz, "in patients with ",Symptom )
paste("Investigating the association between ",Symptom," and ",Parameterz, "in patients with ",Condition )
paste("Investigating the association between ",Symptom," and ",Condition, "in patients with ", Parameterz)
#
# #Patients with condition and a symptom analysis of parameters
# #ie which parameters are significantly elevated in patients with specific symptoms and a specific condition
#
#
# ###### Condition description (ie subclassification)
#
# #1. Clinical and manometric characteristics of patients with oesophagogastric outflow obstruction: towards a new classification George Triadafilopoulos, John O Clarke
#
# ##### Device vs device
# #
# #
# # Both vs future outcomes
# # Both vs future conditions
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.