View source: R/full_threshold.r
FT | R Documentation |
FT begins with a 4-2dB staircase beginning at level
est
. If the final estimate (last seen) is more than 4dB away
from est
, a second 4-2 staircase is completed beginning at the
estimate returned from the first
FT(est = 25, instRange = c(0, 40), verbose = FALSE, makeStim, ...)
FT.start(est = 25, instRange = c(0, 40), makeStim, ...)
FT.step(state, nextStim = NULL)
FT.stop(state)
FT.final(state)
est |
Starting estimate in dB |
instRange |
Dynamic range of the instrument c(min,max) in dB |
verbose |
True if you want each presentation printed |
makeStim |
A function that takes a dB value and numPresentations and returns an OPI datatype ready for passing to opiPresent |
... |
Extra parameters to pass to the opiPresent function |
state |
Current state of the FT returned by |
nextStim |
A valid object for |
This is an implementation of a 4-2 1-up 1-down staircase as
implemented in the first Humphrey Field Analyzer. The initial staircase
starts at est
and proceeds in steps of 4 dB until the first
reversal, and 2dB until the next reversal. The last seen stimulus is
taken as the threshold value. If, after the first staircase, the
threshold is more than 4 dB away from the starting point, then a second
staircase is initiated with a starting point equal to the threshold
found with the first staircase.
Note this function will repeatedly call opiPresent
for a stimulus
until opiPresent
returns NULL
(ie no error occured)
If more than one FT is to be interleaved (for example, testing multiple
locations), then the FT.start
, FT.step
, FT.stop
and FT.final
calls can maintain the state of the FT after each
presentation, and should be used. If only a single FT is required, then
the simpler FT
can be used. See examples below
Returns a list containing
* npres
, total number of presentations.
* respSeq
, response sequence stored as a list of (seen,dB) pairs.
* first
, first staircase estimate in dB.
* final
, final threshold estimate in dB.
FT.start
returns a list that can be passed to FT.step
,
FT.stop
, and FT.final
. It represents the state of a FT
at a single location at a point in time and contains the following.
* name
, FT
.
* startingEstimate=est
, input param.
* currentLevel
, the next stimulus to present.
* minStimulus=instRange[1]
, input param.
* maxStimulus=instRange[2]
, input param.
* makeStim
, input param.
* lastSeen
, the last seen stimulus.
* lastResponse
, the last response given.
* stairResult
, The final result if finished (initially NA
).
* finished
, "Not"
if staircase has not finished, or one of
"Rev"
(finished due to 2 reversals),
"Max"
(finished due to 2 maxStimulus
seen),
"Min"
(finished due to 2 minStimulus
not seen).
* verbose
, number of reversals so far.
* numberOfReversals
, number of reversals so far.
* currSeenLimit
, number of times maxStimulus
has been seen.
* currNotSeenLimit
, number of times minStimulus
not seen.
* numPresentations
, number of presentations so far.
* stimuli
, vector of stimuli shown at each call to FT.step
.
* responses
, vector of responses received (1 seen, 0 not) received at each call to FT.step
.
* responseTimes
, vector of response times received at each call to FT.step
.
* opiParams=list(...)
, input param
* finished
, TRUE
if staircase has finished (2 reversals, or max/min seen/not-seen twice).
FT.step
returns a list containing
state
, the new state after presenting a stimuli and getting a response.
resp
, the return from the opiPresent
call that was made.
FT.stop
returns TRUE
if the first staircase has had 2 reversals, or
maxStimulus
is seen twice or minStimulus
is not seen twice and the
final estimate is within 4 dB of the starting stimulus. Returns TRUE
if
the second staircase has had 2 reversals, or maxStimulus
is seen twice or
minStimulus
is not seen twice
FT.final
returns the final estimate of threshold based on state, which is
the last seen in the second staircase, if it ran, or the first staircase otherwise
FT.final.details
returns a list containing
final
, the final threshold.
first
, the threshold determined by the first staircase (might be different from final).
stopReason
, either Reversals
, Max
, or Min
which are the three ways in which FT can terminate.
np
, number of presentation for the whole procedure (including both staircases if run).
A. Turpin, P.H. Artes and A.M. McKendrick. "The Open Perimetry Interface: An enabling tool for clinical visual psychophysics", Journal of Vision 12(11) 2012.
H. Bebie, F. Fankhauser and J. Spahr. "Static perimetry: strategies", Acta Ophthalmology 54 1976.
C.A. Johnson, B.C. Chauhan, and L.R. Shapiro. "Properties of staircase procedures for estimating thresholds in automated perimetry", Investagative Ophthalmology and Vision Science 33 1993.
dbTocd
, opiPresent
, fourTwo.start
# Stimulus is Size III white-on-white as in the HFA
makeStim <- function(db, n) {
s <- list(x=9, y=9, level=dbTocd(db), size=0.43, color="white",
duration=200, responseWindow=1500)
class(s) <- "opiStaticStimulus"
return(s)
}
chooseOpi("SimHenson")
if (!is.null(opiInitialize(type="C", cap=6)))
stop("opiInitialize failed")
result <- FT(makeStim=makeStim, tt=30, fpr=0.15, fnr=0.01)
if (!is.null(opiClose()))
warning("opiClose() failed")
##############################################
# This section is for multiple FTs
##############################################
makeStimHelper <- function(db,n, x, y) { # returns a function of (db,n)
ff <- function(db, n) db+n
body(ff) <- substitute({
s <- list(x=x, y=y, level=dbTocd(db), size=0.43, color="white",
duration=200, responseWindow=1500)
class(s) <- "opiStaticStimulus"
return(s)
}, list(x=x,y=y))
return(ff)
}
# List of (x, y, true threshold) triples
locations <- list(c(9,9,30), c(-9,-9,32), c(9,-9,31), c(-9,9,33))
# Setup starting states for each location
states <- lapply(locations, function(loc) {
FT.start(makeStim=makeStimHelper(db,n,loc[1],loc[2]),
tt=loc[3], fpr=0.03, fnr=0.01)})
# Loop through until all states are "stop"
while(!all(st <- unlist(lapply(states, FT.stop)))) {
i <- which(!st) # choose a random,
i <- i[runif(1, min=1, max=length(i))] # unstopped state
r <- FT.step(states[[i]]) # step it
states[[i]] <- r$state # update the states
}
finals <- lapply(states, FT.final) # get final estimates of threshold
for(i in 1:length(locations)) {
cat(sprintf("Location (%+2d,%+2d) ",locations[[i]][1], locations[[i]][2]))
cat(sprintf("has threshold %4.2f\n", finals[[i]]))
}
if(!is.null(opiClose()))
warning("opiClose() failed")
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.