In jtrampush/npsych.data: BrainWorkup Neuropsychological Report Writing Package

Data

• Export SLK and/or CSV file from Q-interactive

Setup

# Set JAVA environment
Sys.setenv(
  JAVA_HOME =
    "/Library/Java/JavaVirtualMachines/temurin-11.jdk/Contents/Home"
)
options(java.parameters = "-Xmx16000m")
knitr::opts_chunk$set(
  root.dir = normalizePath("./"),
  echo = FALSE,
  eval = TRUE,
  include = TRUE,
  message = FALSE,
  warning = FALSE,
  error = TRUE
)
library(tidyverse)
library(tabulapdf)
library(rJava)
library(shiny)
library(here)
library(pdftools)
library(fs)
library(magrittr)
library(hablar)
library(googledrive)
library(bwu)

Patient

patient <- params$patient

Test

Name of neuropsych test or rating scale.

test <- params$test
test_name <- params$test_name

Load text/csv file and save as file.path

# input_file_path <- file.choose()
input_file_path <- file.path(params$file)

DKEFS Color-Word Interference Scores from SLK file

library(tidyverse)
test_name_prefix <- "D-KEFS "
output_file_path <- "data/processed_dkefs_cwi_data.csv"

# function
pluck_dkefs_cwi <- function(input_file_path,
                            test_name_prefix,
                            output_file_path = NULL) {
  df <- read_csv(
    input_file_path,
    col_names = FALSE,
    show_col_types = FALSE,
    # locale = locale(encoding = "UTF-16LE"),
    skip = 1
  )

  # Rename the columns
  names(df) <- c("Subtest",
                 "Raw score",
                 "Scaled score",
                 "Percentile rank",
                 "95% CI")

  # Remove the fifth column
  df <- df |> select(Subtest, `Raw score`, `Scaled score`, `Percentile rank`)

  # Find the start of the "CWI" section
  start_line <- which(df == "D-KEFS Color-Word Interference Test") + 1

  # Find the stop of the "CWI" section
  stop_line <- which(df == "Inhibition/Switching Total Errors")

  # Read from the "CWI" section
  df_cwi <- df |>
    dplyr::slice(start_line:stop_line)

  # Keep only rows with the specified prefix in the first column
  # df_raw <- df_raw |> filter(str_starts(Subtest, test_name_prefix))

  # Your new names stored in a character vector (ensure it matches the number of columns in `df`)
  vars <- c("scale", "raw_score", "score", "percentile")

  # Use `set_names()` to rename the columns
  df_cwi <- df_cwi |> set_names(vars)
  df_cwi$scale <- as.character(df_cwi$scale)
  df_cwi$raw_score <- as.numeric(df_cwi$raw_score)
  df_cwi$score <- as.numeric(df_cwi$score)
  df_cwi$percentile <- as.numeric(df_cwi$percentile)

  # Write the combined data to a CSV file
  write_excel_csv(df_cwi, output_file_path, na = "")

  return(df_cwi)

}

df_cwi <- pluck_dkefs_cwi(input_file_path, output_file_path = output_file_path, test_name_prefix)

MUTATE

Create percentile

#TODO: Add percentile calculation ONLY if NA
df_cwi <- df_cwi |> 
  dplyr::mutate(z = (score - 10) / 3) |>
  dplyr::mutate(percentile = ifelse(is.na(percentile), trunc(pnorm(z) * 100), percentile)) |>
  dplyr::select(-z)

# df_cwi <- df_cwi |> 
#   dplyr::mutate(z = (score - 10) / 3) |>
#   dplyr::mutate(percentile = trunc(pnorm(z) * 100)) |>
#   dplyr::select(-z)

Calculate CI 95%

scale_score <- as.numeric(df_cwi$score)
scale_mean <- as.numeric(params$mean)
scale_sd <- as.numeric(params$stdev)
scale_rel <- as.numeric(params$reliability)

ci_values <- bwu::calc_ci_95(
  ability_score = scale_score,
  mean = scale_mean,
  standard_deviation = scale_sd,
  reliability = scale_rel
)
data <- df_cwi
data$true_score <- ci_values["true_score"]
data$ci_lo <- ci_values["lower_ci_95"]
data$ci_hi <- ci_values["upper_ci_95"]
data$ci_95 <- paste0(data$ci_lo, " - ", data$ci_hi)

dkefs <- df_cwi

dkefs <- bwu::gpluck_make_columns(
  data = dkefs,
  test = params$test,
  test_name = params$test_name,
  range = "",
  # ci_95 = paste0(dkefs$ci_95_lower, "-", dkefs$ci_95_upper),
  domain = "",
  subdomain = "",
  narrow = "",
  pass = "",
  verbal = "",
  timed = "",
  test_type = "npsych_test",
  score_type = "scaled_score",
  description = "",
  result = ""
)

Test score ranges

dkefs <- bwu::gpluck_make_score_ranges(
  table = dkefs,
  test_type = "npsych_test"
)

Domain

dkefs <-
  dkefs |> 
  dplyr::mutate(
    domain = dplyr::case_when(
      scale == "Color Naming" ~ "Verbal/Language",
      scale == "Word Reading" ~ "Verbal/Language",
      scale == "Inhibition" ~ "Attention/Executive",
      scale == "Inhibition/Switching" ~ "Attention/Executive",
      scale == "Color Naming Total Errors" ~ "Attention/Executive",
      scale == "Word Reading Total Errors" ~ "Attention/Executive",
      scale == "Inhibition Total Errors" ~ "Attention/Executive",
      scale == "Inhibition/Switching Total Errors" ~ "Attention/Executive",
      TRUE ~ domain
    )
  )

Subdomain

dkefs <-
  dkefs |>
  dplyr::mutate(
    subdomain = dplyr::case_when(
      scale == "Color Naming" ~ "Retrieval",
      scale == "Word Reading" ~ "Retrieval",
      scale == "Inhibition" ~ "Executive Functioning",
      scale == "Inhibition/Switching" ~ "Executive Functioning",
      scale == "Color Naming Total Errors" ~ "Attention",
      scale == "Word Reading Total Errors" ~ "Attention",
      scale == "Inhibition Total Errors" ~ "Attention",
      scale == "Inhibition/Switching Total Errors" ~ "Attention",
      TRUE ~ subdomain
    )
  )

Narrow

dkefs <-
  dkefs |>
  mutate(
    narrow = case_when(
      scale == "Color Naming" ~ "Naming Facility",
      scale == "Word Reading" ~ "Naming Facility",
      scale == "Inhibition" ~ "Inhibition",
      scale == "Inhibition/Switching" ~ "Set-Shifting",
      scale == "Color Naming Total Errors" ~ "Response Monitoring",
      scale == "Word Reading Total Errors" ~ "Response Monitoring",
      scale == "Inhibition Total Errors" ~ "Response Monitoring",
      scale == "Inhibition/Switching Total Errors" ~ "Response Monitoring",
      TRUE ~ narrow
    )
  )

Timed/Untimed

dkefs <-
  dkefs |>
  mutate(
    timed = case_when(
      scale == "Color Naming" ~ "Timed",
      scale == "Word Reading" ~ "Timed",
      scale == "Inhibition" ~ "Timed",
      scale == "Inhibition/Switching" ~ "Timed",
      scale == "Color Naming Total Errors" ~ "",
      scale == "Word Reading Total Errors" ~ "",
      scale == "Inhibition Total Errors" ~ "Timed",
      scale == "Inhibition/Switching Total Errors" ~ "Timed",
      TRUE ~ timed
    )
  )

Verbal/Nonverbal

dkefs <-
  dkefs |>
  mutate(
    verbal = case_when(
      scale == "Color Naming" ~ "Verbal",
      scale == "Word Reading" ~ "Verbal",
      scale == "Inhibition" ~ "Verbal",
      scale == "Inhibition/Switching" ~ "Verbal",
      scale == "Color Naming Total Errors" ~ "Verbal",
      scale == "Word Reading Total Errors" ~ "Verbal",
      scale == "Inhibition Total Errors" ~ "Verbal",
      scale == "Inhibition/Switching Total Errors" ~ "Verbal",
      TRUE ~ verbal
    )
  )

PASS

dkefs <-
  dkefs |>
  mutate(
    pass = case_when(
      scale == "Color Naming" ~ "Sequential",
      scale == "Word Reading" ~ "Sequential",
      scale == "Inhibition" ~ "Attention",
      scale == "Inhibition/Switching" ~ "Attention",
      scale == "Color Naming Total Errors" ~ "Attention",
      scale == "Word Reading Total Errors" ~ "Attention",
      scale == "Inhibition Total Errors" ~ "Attention",
      scale == "Inhibition/Switching Total Errors" ~ "Attention",
      TRUE ~ as.character(pass)
    )
  )

Score type

dkefs <-
  dkefs |>
  mutate(
    score_type = case_when(
      scale == "Color Naming Total Errors" ~ "percentile",
      scale == "Word Reading Total Errors" ~ "percentile",
      TRUE ~ as.character(score_type)
    )
  )

Descriptions

dkefs <-
  dkefs |>
  mutate(
    description = case_when(
      scale == "Color Naming" ~ "Rapid color naming",
      scale == "Word Reading" ~ "Rapid word reading",
      scale == "Inhibition" ~ "Inhibition/cognitive control",
      scale == "Inhibition/Switching" ~ "Set-shifting/cognitive flexibility",
      scale == "Color Naming Total Errors" ~ "Color naming errors",
      scale == "Word Reading Total Errors" ~ "Word reading errors",
      scale == "Inhibition Total Errors" ~ "Inhibition errors",
      scale == "Inhibition/Switching Total Errors" ~ "Switching errors",
      TRUE ~ as.character(description)
    )
  )

Glue results

dkefs <-
  dkefs |>
  dplyr::mutate(
    result = glue::glue("{patient}'s score on {.data$scale} ({.data$description}) was {.data$range}.")
  )

EXPORT

readr::write_csv(dkefs, here::here("data", "dkefs.csv"), col_names = TRUE, na = "")

cat("Finished!")

jtrampush/npsych.data documentation built on Feb. 25, 2025, 12:30 a.m.