teethr

⚠️ VERY early development! ⚠️

The goal of teethr (pronounced teether) is to provide a set of tools to calculate dental indices, such as caries ratio, and calculus index, and facilitate visualisation of data related to dental diseases.

Installation

You can install the development version of teethr like so:

devtools::install_github("bbartholdy/teethr")

Example

First, dental data are combined with demographic data and converted to long format with dental_longer(). dental_longer() also adds information on tooth type and position that can be used to calculate tooth- and position-specific indices.

library(teethr)
library(dplyr)

mb11_calculus_long <- mb11_calculus %>% 
  dental_longer(-id, names_sep = "_") %>% # make longer all columns except 'id'
  inner_join(mb11_demography, by = "id") # combine with age and sex
mb11_calculus_long

#> # A tibble: 3,936 × 6
#>    id    tooth surface score age   sex  
#>    <chr> <chr> <chr>   <dbl> <chr> <chr>
#>  1 MB131 t11   bucc        0 ma    m    
#>  2 MB131 t11   lin         0 ma    m    
#>  3 MB131 t11   ip          1 ma    m    
#>  4 MB131 t12   bucc        0 ma    m    
#>  5 MB131 t12   lin         0 ma    m    
#>  6 MB131 t12   ip          1 ma    m    
#>  7 MB131 t13   bucc        0 ma    m    
#>  8 MB131 t13   lin         0 ma    m    
#>  9 MB131 t13   ip          0 ma    m    
#> 10 MB131 t14   bucc        1 ma    m    
#> # ℹ 3,926 more rows

The long-format data can be used to calculate indices, such as the dental calculus index using the calculus_index() function. The function also provides the number of surfaces scored, number of teeth, and sum of scores for verification. The original method recommends grouping the dentition by quadrant to obtain four indices. This can be done for the whole sample,

mb11_calculus_long %>%
  dental_join() %>%
  group_by(quadrant) %>% 
  calculus_index()
#> Joining with `by = join_by(tooth)`
#> previously defined groups (quadrant) were used. If that was not expected, use
#> `ungroup()` on the data frame before using the function.
#> # A tibble: 4 × 4
#>   quadrant     n score_sum index
#>   <chr>    <int>     <dbl> <dbl>
#> 1 LA         632       631 0.998
#> 2 LP         857       513 0.599
#> 3 UA         554       308 0.556
#> 4 UP         813       541 0.665

or by sex,

mb11_calculus_long %>%
  dental_join() %>%
  # groups can also be added directly to calculus_index()
  calculus_index(sex, quadrant) %>% 
  select(sex, quadrant, index) # remove unnecessary outputs
#> Joining with `by = join_by(tooth)`
#> previously defined groups (sex,quadrant) were used. If that was not expected,
#> use `ungroup()` on the data frame before using the function.
#> # A tibble: 16 × 3
#>    sex   quadrant  index
#>    <chr> <chr>     <dbl>
#>  1 f     LA       0.778 
#>  2 f     LP       0.345 
#>  3 f     UA       0.0556
#>  4 f     UP       0.5   
#>  5 m     LA       1.01  
#>  6 m     LP       0.582 
#>  7 m     UA       0.545 
#>  8 m     UP       0.630 
#>  9 pf    LA       0.667 
#> 10 pf    LP       0.667 
#> 11 pf    UA       0.154 
#> 12 pf    UP       0.688 
#> 13 pm    LA       1.03  
#> 14 pm    LP       0.678 
#> 15 pm    UA       0.671 
#> 16 pm    UP       0.782

or by age and sex.

mb11_calculus_long %>%
  dental_join() %>%
  calculus_index(sex, age, quadrant) %>% 
  select(sex, quadrant, index) # remove unnecessary outputs
#> Joining with `by = join_by(tooth)`
#> previously defined groups (sex,age,quadrant) were used. If that was not
#> expected, use `ungroup()` on the data frame before using the function.
#> Warning in dental_index(., score = {: 1 rows removed because of no teeth
#> present in groupings.
#> # A tibble: 39 × 3
#>    sex   quadrant  index
#>    <chr> <chr>     <dbl>
#>  1 f     LA       0.778 
#>  2 f     LP       0.345 
#>  3 f     UA       0.0556
#>  4 f     UP       0.5   
#>  5 m     LA       0.9   
#>  6 m     LP       0.167 
#>  7 m     UA       0.167 
#>  8 m     UP       0.296 
#>  9 m     LA       0.930 
#> 10 m     LP       0.549 
#> # ℹ 29 more rows

library(ggplot2)
mb11_calculus_long %>% 
  dental_join() %>%
  calculus_index(id, age, quadrant) %>% 
  #filter(quadrant == "UP") %>% 
  ggplot(aes(x = age, y = index, fill = age)) +
    #geom_violin(aes(fill = sex)) +
    geom_boxplot(width = 0.5) +
    scale_fill_viridis_d() +
    facet_wrap(~ quadrant) +
    theme_bw()
#> Joining with `by = join_by(tooth)`
#> previously defined groups (id,age,quadrant) were used. If that was not
#> expected, use `ungroup()` on the data frame before using the function.
#> Warning in dental_index(., score = {: 1 rows removed because of no teeth
#> present in groupings.

Dental caries ratios can be calculated using the caries_ratio function.

mb11_caries_long <- mb11_caries %>% 
  dental_longer(-id) %>%
  dental_join()
#> Joining with `by = join_by(tooth)`

mb11_caries_long %>% 
  count_caries(caries = score, no_lesion = "none") %>%
  group_by(type) %>% 
  dental_ratio(count = caries_count)
#> previously defined groups (type) were used. If that was not expected, use `ungroup()` on the data frame before using the function.
#> # A tibble: 8 × 4
#>   type      n count  ratio
#>   <chr> <int> <dbl>  <dbl>
#> 1 c       140    14 0.1   
#> 2 i1      123     8 0.0650
#> 3 i2      130     7 0.0538
#> 4 m1      117    32 0.274 
#> 5 m2      101    33 0.327 
#> 6 m3       88    23 0.261 
#> 7 pm1     140    14 0.1   
#> 8 pm2     111    18 0.162

The package also facilitates working with the {tidyverse}.

# count number of teeth per individual
mb11_caries_long %>% 
  remove_missing(vars = "score") %>% 
  count(id)
#> Warning: Removed 362 rows containing missing values.
#> # A tibble: 41 × 2
#>    id        n
#>    <chr> <int>
#>  1 MB107    24
#>  2 MB116    20
#>  3 MB117    21
#>  4 MB120    29
#>  5 MB121    27
#>  6 MB131    24
#>  7 MB158    29
#>  8 MB163    24
#>  9 MB18     29
#> 10 MB180    24
#> # ℹ 31 more rows