nv: Normative values generation and management

Description Usage Arguments Value Structure of normative values Examples

Description

Functions to generate and handle normative values. Check section Structure of normative values below for details about how to generate functioning normative values

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
nvgenerate(
  vf,
  method = "pointwise",
  probs = c(0, 0.005, 0.01, 0.02, 0.05, 0.95, 0.98, 0.99, 0.995, 1),
  name = "",
  perimetry = "static automated perimetry",
  strategy = "",
  size = ""
)

agelm(vf)

tddef(agem)

ghdef(perc = 0.85)

pddef(ghfun = ghdef(0.85))

lutdef(vf, probs, type = "quantile", ...)

gdef(agem, sdtd, sdpd)

lutgdef(g, probs, type = "quantile", ...)

Arguments

vf

visual field data with sensitivity values

method

method to generate normative values, pointwise ('pointwise') or smoothed with 2-dimensional quadratic functions ('smooth')

probs

numeric vector of probabilities with values in [0,1]. The values 0 and 1 must be included

name

name for the normative values, e.g., "SUNY-IU pointwise NVs". Default is blank

perimetry

perimetry used to obtain normative data, e.g., "static automated perimetry" (default)

strategy

psychophysical strategy used to obtain threshold values, e.g., "SITA standard". Default is blank

size

stimulus size, if the same size was used for all visual field locations or empty (default)

agem

age model to construct the function to obtain TD values

perc

the percentile to obtain the ranked TD value as reference for the general height (GH) of the visual field. Default is the 85th percentile, thus 0.85

ghfun

function used for determination of the GH and PD values

type

type of estimation for the weighted quantile values. See wtd.quantile for details. Default is 'quantile'

...

arguments to be passed to or from methods

sdtd

standard deviations obtained for TD values

sdpd

standard deviations obtained for PD values

g

a table with global indices

Value

nvgenerate returns a list with normative values

agelm returns a list with coefficients and a function defining a linear age model

tddef returns a function for the computation of TD values

ghdef returns a function for the computation of the general height

pddef returns a function for the computation of PD values

lutdef returns a look up table and a function for the computation of the probability values for TD and PD

gdef returns a function to compute global indices

lutgdef returns a look up table and a function for the computation of the probability values for global indices

Structure of normative values

This is one of the most complex structures in visualFields. It is necessary to be able to run statistical analyses of visual fields obtained from perimetry and it requires data from healthy eyes for its generation. The normative values are only as good as the data they are generated from. Two common ways to generate full normative values from a dataset of healthy eyes, are provided in the package, depending on the method selected. The first one, method="pointwise", generates normative values directly from pointwise statistics. The second one, method="smooth", uses a 2D quadratic functions to smooth out those pointwise statistics. Variations or improvements can be regenerated by copying the code in those functions and editing it.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
# generate normative values from SUNY-IU dataset of healthy eyes
# pointwise
sunyiu_24d2_pw <- nvgenerate(vfctrSunyiu24d2, method = "pointwise",
                             name      = "SUNY-IU pointwise NVs",
                             perimetry = "static automated perimetry",
                             strategy  = "SITA standard",
                             size      = "Size III")
# smooth
sunyiu_24d2 <- nvgenerate(vfctrSunyiu24d2, method = "smooth",
                          name      = "SUNY-IU smoothed NVs",
                          perimetry = "static automated perimetry",
                          strategy  = "SITA standard",
                          size      = "Size III")

visualFields documentation built on Aug. 17, 2021, 1:06 a.m.