R/neurovol.R
In bbuchsbaum/neuroim2: Data Structures and IO for Neuroimaging Data with an Emphasis on Functional Magnetic Resonance Imaging (FMRI)
Defines functions
.pruneCoords
fast.expand.grid
read_vol
NeuroVolSource
LogicalNeuroVol
SparseNeuroVol
DenseNeuroVol
NeuroVol
Documented in
DenseNeuroVol
LogicalNeuroVol
NeuroVol
NeuroVolSource
read_vol
SparseNeuroVol
#' @importFrom assertthat assert_that
#' @importFrom Matrix sparseVector
#' @include all_class.R
#' @include all_generic.R
NULL
#' Construct a NeuroVol object
#'
#' Construct a \code{\linkS4class{NeuroVol}} instance, using default (dense) implementation
#' @param data a three-dimensional \code{array}
#' @param space an instance of class \code{\linkS4class{NeuroSpace}}
#' @param label a \code{character} string to identify volume
#' @param indices an 1D vector that gives the linear indices of the associated \code{data} vector
#' @return a \code{\linkS4class{DenseNeuroVol}} instance
#' @examples
#' bspace <- NeuroSpace(c(64,64,64), spacing=c(1,1,1))
#' dat <- array(rnorm(64*64*64), c(64,64,64))
#' bvol <- NeuroVol(dat,bspace, label="test")
#' @export NeuroVol
#' @rdname NeuroVol
NeuroVol <- function(data, space, label="", indices=NULL) {
DenseNeuroVol(data,space, label=label, indices=indices)
}
#' Construct a DenseNeuroVol object
#'
#' Construct a \code{\linkS4class{DenseNeuroVol}} instance
#' @param data a three-dimensional \code{array}
#' @param space an instance of class \code{\linkS4class{NeuroSpace}}
#' @param label a \code{character} string
#' @param indices an optional 1-d index vector
#' @return \code{\linkS4class{DenseNeuroVol}} instance
#' @export DenseNeuroVol
#' @rdname DenseNeuroVol-class
DenseNeuroVol <- function(data, space, label="", indices=NULL) {
if (length(dim(space)) != 3) {
stop("DenseNeuroVol: space argument must have three dimensions")
}
if (is.matrix(data)) {
if (nrow(data) == 1 || ncol(data) == 1) {
data <- as.vector(data)
}
}
if (length(data) == prod(dim(space)) && is.vector(data)) {
dim(data) <- dim(space)
}
if (ndim(space) != 3) {
stop("DenseNeuroVol: space argument must have three dimensions")
}
if (!all(dim(space) == dim(data))) {
stop("DenseNeuroVol: data and space argument must have equal dimensions")
}
if (!is.null(indices)) {
assert_that(length(indices) == length(data))
newdat <- array(0, dim(space))
newdat[indices] <- data
data <- newdat
}
new("DenseNeuroVol", data, space=space)
}
#' SparseNeuroVol
#'
#' Construct a \code{\linkS4class{SparseNeuroVol}} instance
#' @param data a numeric vector
#' @param space an instance of class \code{\linkS4class{NeuroSpace}}
#' @param indices a index vector indicating the 1-d coordinates of the data values
#' @param label a \code{character} string
#' @return \code{\linkS4class{SparseNeuroVol}} instance
#' @export SparseNeuroVol
#' @import Matrix
#' @details
#' Image data is backed by \code{Matrix::sparseVector}.
#' @examples
#' data <- 1:10
#' indices <- seq(1,1000, length.out=10)
#' bspace <- NeuroSpace(c(64,64,64), spacing=c(1,1,1))
#' sparsevol <- SparseNeuroVol(data,bspace,indices=indices)
#' densevol <- NeuroVol(data,bspace,indices=indices)
#' sum(sparsevol) == sum(densevol)
#'
#'
#'
#' @rdname SparseNeuroVol-class
SparseNeuroVol <- function(data, space, indices=NULL, label="") {
if (length(indices) != length(data)) {
stop(paste("length of 'data' must equal length of 'indices'"))
}
sv <- Matrix::sparseVector(x=data, i=indices, length=prod(dim(space)))
new("SparseNeuroVol", data=sv, space=space)
}
#' LogicalNeuroVol
#'
#' This function constructs a \code{\linkS4class{LogicalNeuroVol}} instance.
#'
#' @param data A three-dimensional \code{array}, a 1D vector with length equal to \code{prod(dim(space))}, or a set of \code{indices} where elements are \code{TRUE}.
#' @param space An instance of class \code{\linkS4class{NeuroSpace}}.
#' @param label A \code{character} string.
#' @param indices An optional 1-d index vector.
#'
#' @return A \code{\linkS4class{LogicalNeuroVol}} instance.
#'
#' @examples
#' # Load an example brain mask
#' brain_mask <- read_vol(system.file("extdata", "global_mask.nii", package="neuroim2"))
#'
#' # Convert the brain mask to a LogicalNeuroVol
#' logical_vol <- LogicalNeuroVol(brain_mask, space(brain_mask))
#'
#' @export LogicalNeuroVol
#' @rdname LogicalNeuroVol-class
LogicalNeuroVol <- function(data, space, label="", indices=NULL) {
if (is.null(dim(data)) && length(data) == prod(dim(space))) {
data <- array(data, dim(space))
} else if (is.null(dim(data)) && !is.null(indices)) {
newdat <- array(FALSE, dim(space))
newdat[indices] <- data
data <- newdat
}
if (length(dim(data)) != 3) {
stop("LogicalNeuroVol: data argument must have three dimensions")
}
if (ndim(space) != 3) {
stop("LogicalVolume: space argument must have three dimensions")
}
if (!is.logical(data)) {
D <- dim(data)
data <- as.logical(data > 0)
dim(data) <- D
}
new("LogicalNeuroVol", .Data=data, space=space)
}
#' Convert DenseNeuroVol to array
#'
#' This function converts a DenseNeuroVol object to an array.
#'
#' @param from A DenseNeuroVol object.
#'
#' @return An array resulting from the conversion.
#'
#' @keywords internal
#' @name as,DenseNeuroVol,array
setAs(from="DenseNeuroVol", to="array", def=function(from) from@.Data)
#' Convert DenseNeuroVol to H5NeuroVol
#'
#' This function converts a DenseNeuroVol object to an H5NeuroVol object.
#'
#' @param from A DenseNeuroVol object.
#'
#' @return An H5NeuroVol object resulting from the conversion.
#'
#' @keywords internal
#' @name DenseNeuroVol,H5NeuroVol
setAs(from="DenseNeuroVol", to="H5NeuroVol", def=function(from) {
to_nih5_vol(from, file_name=NULL, data_type="FLOAT")
})
#' Convert DenseNeuroVec to H5NeuroVec
#'
#' This function converts a DenseNeuroVec object to an H5NeuroVec object.
#'
#' @param from A DenseNeuroVec object.
#'
#' @return An H5NeuroVec object resulting from the conversion.
#'
#' @keywords internal
#' @name DenseNeuroVec,H5NeuroVec
setAs(from="DenseNeuroVec", to="H5NeuroVec", def=function(from) {
to_nih5_vec(from, file_name=NULL, data_type="FLOAT")
})
#' Convert SparseNeuroVol to array
#'
#' This function converts a SparseNeuroVol object to an array object.
#'
#' @param from A SparseNeuroVol object.
#'
#' @return An array object resulting from the conversion.
#'
#' @keywords internal
#' @name SparseNeuroVol,array
setAs(from="SparseNeuroVol", to="array", def=function(from) {
vals <- as.numeric(from@data)
array(vals, dim(from))
})
#' Convert SparseNeuroVol to numeric
#'
#' This function converts a SparseNeuroVol object to a numeric object.
#'
#' @param from A SparseNeuroVol object.
#'
#' @return A numeric object resulting from the conversion.
#' @keywords internal
#' @name SparseNeuroVol,numeric
setAs(from="SparseNeuroVol", to="numeric", def=function(from) {
as.numeric(from@data)
})
#' Convert SparseNeuroVol to numeric
#'
#' @rdname as.numeric-methods
#' @param x the object to convert
#' @export
setMethod(f="as.numeric", signature=signature(x = "SparseNeuroVol"), def=function(x) {
as(x, "numeric")
})
#' conversion from \code{\linkS4class{NeuroVol}} to \code{\linkS4class{LogicalNeuroVol}}
#'
#' @name as
#' @export
setAs(from="NeuroVol", to="LogicalNeuroVol", def=function(from) {
LogicalNeuroVol(as.array(from), space(from))
})
#' conversion from \code{\linkS4class{NeuroVol}} to \code{array}
#'
#' @name as
#' @export
setAs(from="NeuroVol", to="array", def=function(from) from[,,])
#' show a \code{NeuroVol}
#' @param object the object
#' @export
setMethod(f="show", signature=signature("NeuroVol"),
def=function(object) {
sp <- space(object)
cat("NeuroVol\n")
cat(" Type :", class(object), "\n")
cat(" Dimension :", dim(object), "\n")
cat(" Spacing :", paste(paste(signif(sp@spacing[1:(length(sp@spacing)-1)],2), " X ", collapse=" "),
sp@spacing[length(sp@spacing)], "\n"))
cat(" Origin :", paste(paste(signif(sp@origin[1:(length(sp@origin)-1)],2), " X ", collapse=" "),
sp@origin[length(sp@origin)], "\n"))
cat(" Axes :", paste(sp@axes@i@axis, sp@axes@j@axis,sp@axes@k@axis), "\n")
}
)
#' show a \code{SparseNeuroVol}
#' @param object the object
#' @export
setMethod(f="show", signature=signature("SparseNeuroVol"),
def=function(object) {
sp <- space(object)
cat("NeuroVol\n")
cat(" Type :", class(object), "\n")
cat(" Dimension :", dim(object), "\n")
cat(" Spacing :", paste(paste(signif(sp@spacing[1:(length(sp@spacing)-1)],2), " X ", collapse=" "),
sp@spacing[length(sp@spacing)], "\n"))
cat(" Origin :", paste(paste(signif(sp@origin[1:(length(sp@origin)-1)],2), " X ", collapse=" "),
sp@origin[length(sp@origin)], "\n"))
cat(" Axes :", paste(sp@axes@i@axis, sp@axes@j@axis,sp@axes@k@axis), "\n")
cat(" Cardinality :", length(Matrix::which(object@data>0)))
}
)
#' load a NeuroVol
#' @noRd
## TODO reduce code duplication with load_data#NeuroVecSource
setMethod(f="load_data", signature=c(x="NeuroVolSource"),
def=function(x) {
meta <- x@meta_info
nels <- prod(meta@dims[1:3])
### for brain buckets, this offset needs to be precomputed ....
offset <- (nels * (x@index-1)) * meta@bytes_per_element
reader <- data_reader(meta, offset)
dat <- read_elements(reader, nels)
## bit of a hack to deal with scale factors
if (.hasSlot(meta, "slope")) {
if (meta@slope != 0) {
dat <- dat*meta@slope
}
}
close(reader)
arr <- array(dat, meta@dims[1:3])
bspace <- NeuroSpace(meta@dims[1:3], meta@spacing, meta@origin, meta@spatial_axes, trans(meta))
DenseNeuroVol(arr, bspace, x)
})
#' Constructor for NeuroVolSource
#'
#' @param input the input file name
#' @param index the image subvolume index
#' @export
#' @rdname NeuroVolSource-class
#' @return a new instance of type \code{NeuroVolSource}
NeuroVolSource <- function(input, index=1) {
stopifnot(index >= 1)
stopifnot(is.character(input))
if (!file.exists(input)) {
candidates <- Sys.glob(paste(input, "*", sep=""))
if (length(candidates) > 0) {
input <- candidates[1]
}
}
stopifnot(file.exists(input))
meta_info <- read_header(input)
if ( length(meta_info@dims) < 4 && index > 1) {
stop("index cannot be greater than 1 for a image with dimensions < 4")
}
new("NeuroVolSource", meta_info=meta_info, index=as.integer(index))
}
#' Load an image volume from a file
#'
#' @param file_name the name of the file to load
#' @param index the index of the volume (e.g. if the file is 4-dimensional)
#' @return an instance of the class \code{\linkS4class{DenseNeuroVol}}
#'
#' @examples
#' fname <- system.file("extdata", "global_mask.nii", package="neuroim2")
#' x <- read_vol(fname)
#' print(dim(x))
#' space(x)
#'
#' @export read_vol
read_vol <- function(file_name, index=1) {
src <- NeuroVolSource(file_name, index)
load_data(src)
}
#' @export
#' @rdname slices-methods
setMethod(f="slices", signature=signature(x="NeuroVol"),
def = function(x) {
nslices <- dim(x)[3]
f <- function(i) slice(x, i, 3)
#lis <- lapply(1:nslices, function(i) f)
#deferred_list(lis)
deflist::deflist(f, nslices)
})
#' @export
#' @rdname extractor3d
setMethod(f="[", signature=signature(x = "NeuroVol", i = "ROICoords", j = "missing"),
def=function (x, i, j, k, ..., drop=TRUE) {
callGeneric(x, i@coords)
}
)
#' @rdname concat-methods
#' @export
setMethod(f="concat", signature=signature(x="DenseNeuroVol", y="missing"),
def=function(x,y,...) {
x
})
#' @note dimensions of x and y must be equal
#' @export
#' @rdname concat-methods
setMethod(f="concat", signature=signature(x="DenseNeuroVol", y="DenseNeuroVol"),
def=function(x,y,...) {
.concat4D(x,y,...)
})
#' @export
#' @rdname map_values-methods
setMethod(f="map_values", signature=signature(x="NeuroVol", lookup="list"),
def=function(x,lookup) {
if (is.null(names(lookup))) {
names(lookup) <- seq_along(lookup)
}
out <- match(x,as.numeric(names(lookup)))
DenseNeuroVol(unlist(lookup[out]), space(x))
})
#' @export
#' @rdname map_values-methods
setMethod(f="map_values", signature=signature(x="NeuroVol", lookup="matrix"),
def=function(x,lookup) {
if (ncol(lookup) != 2) {
stop("map_values: lookup matrix must have two columns: column 1 is key, column 2 is value")
} else if (nrow(lookup) < 1) {
stop("map_values: lookup matrix have at least one row")
}
m <- match(as.vector(x), lookup[,1])
outv <- lookup[m,2]
outv[is.na(outv)] <- 0
NeuroVol(array(outv, dim(x)), space(x))
})
#' @export
#' @rdname split_fill-methods
setMethod(f="split_fill", signature=signature(x="NeuroVol", fac="factor", FUN="function"),
def=function(x,fac,FUN) {
stopifnot(length(x) == length(fac))
S <- split(1:length(x), fac, drop=TRUE)
res <- map(S, function(ind) {
X <- FUN(x[ind])
if (length(X) == length(ind)) {
X
} else {
rep(X, length(ind))
}
})
ovol <- x
ovol[1:length(x)] <- unsplit(res, fac)
ovol
})
#' @keywords internal
#' @noRd
fast.expand.grid <- function(seq1,seq2, constant) {
cbind(Var1 = rep.int(seq1, length(seq2)),
Var2 = rep.int(seq2, rep.int(length(seq1),length(seq2))),
Var3=constant)
}
#' @export
#' @rdname slice-methods
setMethod(f="slice", signature=signature(x="NeuroVol", zlevel="numeric", along="numeric",
orientation="missing"),
def=function(x, zlevel, along, orientation) {
stopifnot(along >= 1 && along <=3)
stopifnot(zlevel >= 1 && zlevel <= dim(x)[along])
imslice <- switch(as.character(along),
"1"=x[zlevel,,],
"2"=x[,zlevel,],
"3"=x[,,zlevel])
NeuroSlice(imslice, drop_dim(space(x), along))
})
#' @export
#' @rdname slice-methods
setMethod(f="slice", signature=signature(x="NeuroVol", zlevel="numeric", along="NeuroSpace",
orientation="AxisSet3D"),
def=function(x, zlevel, along, orientation) {
xdim <- dim_of(along, orientation@i)
ydim <- dim_of(along, orientation@j)
zdim <- dim_of(along, orientation@k)
#browser()
stopifnot(zlevel >= 1 && zlevel <= zdim)
vox <- as.matrix(fast.expand.grid(seq(1,xdim), seq(1,ydim), zlevel))
gg <- grid_to_grid(along, vox)
imslice <- x[gg]
NeuroSlice(matrix(imslice, xdim,ydim), drop_dim(along,which_dim(along, orientation@k)))
})
#' @title Convert NeuroVol to a mask
#' @description This method converts a NeuroVol object to a mask by setting all positive values to TRUE and all non-positive values to FALSE.
#' @param x A NeuroVol object to convert to a mask.
#' @param indices A missing argument; not used in this method.
#' @return A LogicalNeuroVol object representing the mask created from the input NeuroVol.
#' @export
#' @rdname as.mask-methods
setMethod(f="as.mask", signature=signature(x="NeuroVol", indices="missing"),
def=function(x) {
LogicalNeuroVol(x > 0, space(x))
})
#' @title Convert NeuroVol to a mask with specified indices
#' @description This method converts a NeuroVol object to a mask by setting the specified indices to TRUE and the remaining elements to FALSE.
#' @param x A NeuroVol object to convert to a mask.
#' @param indices A numeric vector containing the indices of the input NeuroVol that should be set to TRUE in the resulting mask.
#' @return A LogicalNeuroVol object representing the mask created from the input NeuroVol with specified indices.
#' @export
#' @rdname as.mask-methods
setMethod(f="as.mask", signature=signature(x="NeuroVol", indices="numeric"),
def=function(x, indices) {
M <- array(0, dim(x))
M[indices] <- 1
LogicalNeuroVol(M, space(x))
})
#' @export
#' @rdname coord_to_grid-methods
setMethod(f="coord_to_grid", signature=signature(x="NeuroVol", coords="matrix"),
def=function(x, coords) {
callGeneric(space(x), coords)
})
#' @export
#' @rdname coord_to_grid-methods
setMethod(f="coord_to_grid", signature=signature(x="NeuroVol", coords="numeric"),
def=function(x, coords) {
coords <- matrix(coords, nrow=1)
callGeneric(x, coords)
})
#' @importFrom dbscan kNN
#' @keywords internal
#' @noRd
.pruneCoords <- function(coord.set, vals, mindist=10) {
if (NROW(coord.set) == 1) {
1
}
.prune <- function(keepIndices) {
if (length(keepIndices) <= 2) {
keepIndices
} else {
ret <- try(dbscan::kNN(coord.set[keepIndices,], coord.set[keepIndices,], k=2))
#ret <- rflann::Neighbour(coord.set[keepIndices,], coord.set[keepIndices,], k=2, build="kdtree", cores=0, checks=1)
#ind <- ret$indices[, 2]
#ds <- sqrt(ret$distances[, 2])
ind <- ret$id[,2]
ds <- ret$distances[,2]
v <- vals[keepIndices]
ovals <- v[ind]
pruneSet <- ifelse(ds < mindist & ovals > v, TRUE, FALSE)
if (any(pruneSet)) {
Recall(keepIndices[!pruneSet])
} else {
keepIndices
}
}
}
.prune(1:NROW(coord.set))
}
#' @title Create a patch set from a NeuroVol object
#' @description This function creates a patch set from a NeuroVol object given specified dimensions
#' @return A deferred list of patches.
#' @param x a NeuroVol object
#' @param dims the dimensions of the patch
#' @param ... additional args
#' @export
#' @rdname patch_set-methods
setMethod(f="patch_set", signature=signature(x="NeuroVol",
dims="numeric",
mask="missing"),
def=function(x, dims, ...) {
mask <- LogicalNeuroVol(array(1, dim(x)), space=space(x))
callGeneric(x, dims, mask)
})
#' @title Create a patch set from a NeuroVol object
#' @description This function creates a patch set from a NeuroVol object given specified dimensions and a mask.
#' @return A deferred list of patches.
#' @param x a NeuroVol object
#' @param dims the dimensions of the patch
#' @param mask the mask defining the valid patch centers
#' @param ... additional args
#' @export
#' @rdname patch_set-methods
setMethod(f="patch_set", signature=signature(x="NeuroVol",
dims="numeric",
mask="LogicalNeuroVol"),
def=function(x, dims, mask, ...) {
if (!all(dims > 0)) {
stop("all 'dims' must be greater than zero")
}
if (!all(dims %% 2 == 1)) {
stop("all 'dims' must be odd numbers")
}
template <- as.matrix(expand.grid
(x=seq(seq(ceiling(-dims[1]/2),floor(dims[1]/2))),
y=seq(seq(ceiling(-dims[2]/2),floor(dims[2]/2))),
z=seq(seq(ceiling(-dims[3]/2),floor(dims[3]/2)))))
xdim <- dim(x)
grid <- index_to_grid(mask, which(mask>0))
f <- function(i) {
g <- grid[i,]
m <- t(t(template) + g)
m[,1] <- pmax(pmin(m[,1], xdim[1]), 1)
m[,2] <- pmax(pmin(m[,2], xdim[2]), 1)
m[,3] <- pmax(pmin(m[,3], xdim[3]), 1)
ret <- x[m]
attr(ret, "idx") <- grid_to_index(x, m)
ret
}
patches <- deflist::deflist(f, nrow(grid))
patches
#patches <- deferred_list(lapply(1:nrow(grid), function(i) f))
})
#' apply a kernel function to a \code{\linkS4class{NeuroVol}}
#'
#' @rdname map-methods
#' @param mask restrict application of kernel to masked area
#' @export
setMethod(f="mapf", signature=signature(x="NeuroVol", m="Kernel"),
def=function(x, m, mask=NULL) {
ovol <- array(0, dim(x))
hwidth <- map_dbl(m@width, function(d) ceiling(d/2 -1)) + 1
xdim <- dim(x)[1]
ydim <- dim(x)[2]
zdim <- dim(x)[3]
if (!is.null(mask)) {
if (!all.equal(dim(mask), dim(ovol))) {
stop(paste("mask must have same dimensions as input volume"))
}
grid <- index_to_grid(mask, which(mask != 0))
} else {
grid <- as.matrix(expand.grid(i=hwidth[1]:(xdim - hwidth[1]),
j=hwidth[2]:(ydim - hwidth[2]),
k=hwidth[3]:(zdim - hwidth[3])))
}
res <- apply(grid, 1, function(vox) {
loc <- t(t(m@voxels) +vox)
ivals <- x[loc]
sum(ivals * m@weights)
})
ovol[grid] <- res
NeuroVol(ovol, space(x))
})
#' find connected components in NeuroVol
#'
#' @export
#' @importFrom purrr map_int
#' @param threshold threshold defining lower intensity bound for image mask
#' @param cluster_table return cluster_table
#' @param local_maxima return table of local maxima
#' @param local_maxima_dist the distance used to define minum distance between local maxima
#' @rdname conn_comp-methods
setMethod(f="conn_comp", signature=signature(x="NeuroVol"),
def=function(x, threshold=0, cluster_table=TRUE, local_maxima=TRUE, local_maxima_dist=15,...) {
mask <- (x > threshold)
stopifnot(any(mask))
comps <- conn_comp_3D(mask,...)
grid <- as.data.frame(index_to_grid(mask, which(mask>0)))
colnames(grid) <- c("x", "y", "z")
locations <- split(grid, comps$index[comps$index>0])
ret <- list(index=ClusteredNeuroVol(mask, clusters=comps$index[mask>0]), size=NeuroVol(comps$size, space(x)), voxels=locations)
if (cluster_table) {
maxima <- do.call(rbind, lapply(locations, function(loc) {
if (nrow(loc) == 1) {
loc
} else {
vals <- x[as.matrix(loc)]
loc[which.max(vals),]
}
}))
N <- comps$size[as.matrix(maxima)]
Area <- N * prod(spacing(x))
maxvals <- x[as.matrix(maxima)]
ret$cluster_table <- data.frame(index=1:NROW(maxima),
x=maxima[,1], y=maxima[,2], z=maxima[,3],
N=N, Area=Area, value=maxvals)
}
if (local_maxima) {
#if (all(map_int(locations, NROW) == 1)) {
#
#}
coord.sets <- lapply(locations, function(loc) {
sweep(as.matrix(loc), 2, spacing(x), "*")
})
loc.max <- do.call(rbind, mapply(function(cset, i) {
idx <- .pruneCoords(as.matrix(cset), x[as.matrix(locations[[i]])], mindist=local_maxima_dist)
maxvox <- as.matrix(locations[[i]])[idx,,drop=F]
cbind(i, maxvox)
}, coord.sets, 1:length(coord.sets), SIMPLIFY=FALSE))
loc.max <- cbind(loc.max, x[loc.max[, 2:4, drop=F]])
row.names(loc.max) <- 1:NROW(loc.max)
colnames(loc.max) <- c("index", "x", "y", "z", "value")
ret$local_maxima <- loc.max
}
ret
})
### TODO when the source voulme is an AFNI BRIK the output file does not preserve orientation
### e.g. write_vol(x, space(BRIK), ...)
### if BRIK was RAI, output NIFTI file reverts to LPI
#' @export
#' @rdname write_vol-methods
setMethod(f="write_vol",signature=signature(x="NeuroVol", file_name="character", format="missing", data_type="missing"),
def=function(x, file_name) {
write_nifti_volume(x, file_name)
})
#' @export
#' @rdname write_vol-methods
setMethod(f="write_vol",signature=signature(x="ClusteredNeuroVol", file_name="character", format="missing", data_type="missing"),
def=function(x, file_name) {
callGeneric(as(x, "DenseNeuroVol"), file_name)
})
#' @export
#' @rdname write_vol-methods
setMethod(f="write_vol",signature=signature(x="NeuroVol", file_name="character", format="character", data_type="missing"),
def=function(x, file_name, format) {
if (toupper(format) == "NIFTI" || toupper(format) == "NIFTI1" || toupper(format) == "NIFTI-1") {
callGeneric(x, file_name)
} else {
stop(paste("sorry, cannot write format: ", format))
}
})
#' @export
#' @rdname write_vol-methods
setMethod(f="write_vol",signature=signature(x="ROIVol", file_name="character", format="character", data_type="missing"),
def=function(x, file_name, format) {
if (toupper(format) == "NIFTI" || toupper(format) == "NIFTI1" || toupper(format) == "NIFTI-1") {
callGeneric(as.dense(x), file_name)
} else {
stop(paste("sorry, cannot write format: ", format))
}
})
#' @export write_vol
#' @rdname write_vol-methods
setMethod(f="write_vol",signature=signature(x="NeuroVol", file_name="character", format="missing", data_type="character"),
def=function(x, file_name, data_type) {
write_nifti_volume(x, file_name, data_type)
})
#' as.logical
#'
#' Convert NeuroVol to \code{\linkS4class{LogicalNeuroVol}}
#'
#' the image values will be converted to using R base function \code{as.logical} and wrapped in \code{LogicalNeuroVol}
#'
#' @param x the object
#' @return an instance of \code{\linkS4class{LogicalNeuroVol}}
#' @rdname as.logical-methods
#' @export
setMethod(f="as.logical", signature=signature(x = "NeuroVol"), def=function(x) {
vals <- as.logical(as.vector(x))
LogicalNeuroVol(vals, space(x))
})
#' @rdname as.sparse-methods
setMethod(f="as.sparse", signature=signature(x="DenseNeuroVol", mask="LogicalNeuroVol"),
def=function(x, mask) {
assert_that(all(dim(x) == dim(mask)))
assert_that(all(spacing(x) == spacing(mask)))
dat <- x[mask]
bvec <- SparseNeuroVol(data=dat, space=space(x),indices=which(mask>0))
})
#' @rdname partition-methods
setMethod("partition", signature=signature(x="LogicalNeuroVol", k="integer"),
def=function(x,k) {
assert_that(k>1)
idx <- which(x != 0)
cds <- index_to_coord(x,idx)
kres <- kmeans(cds, centers=k, iter.max=200)
ClusteredNeuroVol(x, kres$cluster)
})
#' @rdname partition-methods
setMethod("partition", signature=signature(x="LogicalNeuroVol", k="numeric"),
def=function(x,k) {
callGeneric(x, as.integer(k))
})
#' @rdname partition-methods
setMethod("partition", signature=signature(x="DenseNeuroVol", k="numeric"),
def=function(x,k) {
callGeneric(as.logical(x), as.integer(k))
})
#' @rdname as.sparse-methods
setMethod(f="as.sparse", signature=signature(x="DenseNeuroVol", mask="numeric"),
def=function(x, mask) {
m <- as.integer(mask)
bvol <- SparseNeuroVol(x[m], space(x), indices=m)
bvol
})
#' @noRd
setMethod(f="linear_access", signature=signature(x = "SparseNeuroVol", i = "numeric"),
def=function (x, i) {
x@data[as.numeric(i)]
})
#' extractor
#' @export
#' @param x the object
#' @param i first index
#' @param j second index
#' @param k third index
#' @param ... additional args
#' @param drop dimension
setMethod(f="[", signature=signature(x = "SparseNeuroVol", i = "numeric", j = "numeric", drop="ANY"),
def=function (x, i, j, k, ..., drop=TRUE) {
if (missing(k)) {
k <- seq(1, dim(x)[3])
}
nel <- length(i) * length(j) * length(k)
mind <- cbind(rep(i, length.out=nel), rep(j, each=length(i)), rep(k, each=length(i) * length(j)))
ind <- grid_to_index(x, mind)
dat <- x@data[as.numeric(ind)]
array(dat, c(length(i), length(j), length(k)))
})
#' plot an NeuroVol as a series of 2D slices
#'
#' @rdname plot-methods
#' @param x the object to display
#' @param cmap a color map consisting of a vector of colors in hex format (e.g. \code{gray(n=255)})
#' @param zlevels the series of slice indices to display.
#' @param irange the intensity range indicating the low and high values of the color scale.
#' @param thresh a 2-element vector indicating the lower and upper transparency thresholds.
#' @param alpha the level of alpha transparency
#' @param bgvol a background volume that serves as an image underlay (currently ignored).
#' @param bgcmap a color map for backround layer consisting of a vector of colors in hex format (e.g. \code{gray(n=255)})
#' @export
#' @importFrom graphics plot
#' @import colorplane
#' @examples
#'
#' dat <- matrix(rnorm(100*100), 100, 100)
#' slice <- NeuroSlice(dat, NeuroSpace(c(100,100)))
#' #plot(slice)
setMethod("plot", signature=signature(x="NeuroVol"),
def=function(x,cmap=gray(seq(0,1,length.out=255)),
zlevels=unique(round(seq(1, dim(x)[3], length.out=6))),
irange=range(x, na.rm=TRUE),
thresh=c(0,0),
alpha=1,
bgvol=NULL,
bgcmap=gray(seq(0,1,length.out=255))) {
if (!requireNamespace("ggplot2", quietly = TRUE)) {
stop("Package \"ggplot2\" needed for this function to work. Please install it.",
call. = FALSE)
}
if (!is.null(bgvol)) {
assert_that(all(dim(x) == dim(bgvol)))
assert_that(all(spacing(x) == spacing(bgvol)))
}
df1 <- do.call(rbind, purrr::map(zlevels, function(i) {
if (!is.null(bgvol)) {
bgslice <- slice(bgvol, zlevel=i, along=3)
bgplane <- colorplane::IntensityColorPlane(as.numeric(bgslice), cols=bgcmap)
bgcols <- colorplane::map_colors(bgplane)
}
imslice <- slice(x, zlevel=i, along=3)
implane <- colorplane::IntensityColorPlane(as.numeric(imslice), cols=cmap, alpha=alpha)
fgcols <- colorplane::map_colors(implane, threshold=thresh, irange=irange)
if (!is.null(bgvol)) {
fgcols <- colorplane::as_hexcol(colorplane::blend_colors(bgcols, fgcols, alpha=alpha))
} else {
fgcols <- colorplane::as_hexcol(fgcols)
}
cds <- index_to_coord(space(imslice), 1:length(imslice))
data.frame(x=cds[,1], y=cds[,2], z=i, value=as.vector(fgcols))
}))
{y=value=NULL}
p <-
ggplot2::ggplot(ggplot2::aes(x=x, y=y), data=df1) +
ggplot2::coord_fixed() +
ggplot2::geom_raster(ggplot2::aes(x=x,y=y, fill=value)) +
ggplot2::scale_fill_identity() + ggplot2::xlab("") + ggplot2::ylab("") +
ggplot2::scale_x_continuous(expand=c(0,0)) +
ggplot2::scale_y_continuous(expand=c(0,0)) +
ggplot2::facet_wrap(~ z) +
ggplot2::theme_classic()
p
})
bbuchsbaum/neuroim2 documentation built on March 29, 2024, 6:36 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 .pruneCoords fast.expand.grid read_vol NeuroVolSource LogicalNeuroVol SparseNeuroVol DenseNeuroVol NeuroVol
Documented in DenseNeuroVol LogicalNeuroVol NeuroVol NeuroVolSource read_vol SparseNeuroVol
#' @importFrom assertthat assert_that
#' @importFrom Matrix sparseVector
#' @include all_class.R
#' @include all_generic.R
NULL
#' Construct a NeuroVol object
#'
#' Construct a \code{\linkS4class{NeuroVol}} instance, using default (dense) implementation
#' @param data a three-dimensional \code{array}
#' @param space an instance of class \code{\linkS4class{NeuroSpace}}
#' @param label a \code{character} string to identify volume
#' @param indices an 1D vector that gives the linear indices of the associated \code{data} vector
#' @return a \code{\linkS4class{DenseNeuroVol}} instance
#' @examples
#' bspace <- NeuroSpace(c(64,64,64), spacing=c(1,1,1))
#' dat <- array(rnorm(64*64*64), c(64,64,64))
#' bvol <- NeuroVol(dat,bspace, label="test")
#' @export NeuroVol
#' @rdname NeuroVol
NeuroVol <- function(data, space, label="", indices=NULL) {
DenseNeuroVol(data,space, label=label, indices=indices)
}
#' Construct a DenseNeuroVol object
#'
#' Construct a \code{\linkS4class{DenseNeuroVol}} instance
#' @param data a three-dimensional \code{array}
#' @param space an instance of class \code{\linkS4class{NeuroSpace}}
#' @param label a \code{character} string
#' @param indices an optional 1-d index vector
#' @return \code{\linkS4class{DenseNeuroVol}} instance
#' @export DenseNeuroVol
#' @rdname DenseNeuroVol-class
DenseNeuroVol <- function(data, space, label="", indices=NULL) {
if (length(dim(space)) != 3) {
stop("DenseNeuroVol: space argument must have three dimensions")
}
if (is.matrix(data)) {
if (nrow(data) == 1 || ncol(data) == 1) {
data <- as.vector(data)
}
}
if (length(data) == prod(dim(space)) && is.vector(data)) {
dim(data) <- dim(space)
}
if (ndim(space) != 3) {
stop("DenseNeuroVol: space argument must have three dimensions")
}
if (!all(dim(space) == dim(data))) {
stop("DenseNeuroVol: data and space argument must have equal dimensions")
}
if (!is.null(indices)) {
assert_that(length(indices) == length(data))
newdat <- array(0, dim(space))
newdat[indices] <- data
data <- newdat
}
new("DenseNeuroVol", data, space=space)
}
#' SparseNeuroVol
#'
#' Construct a \code{\linkS4class{SparseNeuroVol}} instance
#' @param data a numeric vector
#' @param space an instance of class \code{\linkS4class{NeuroSpace}}
#' @param indices a index vector indicating the 1-d coordinates of the data values
#' @param label a \code{character} string
#' @return \code{\linkS4class{SparseNeuroVol}} instance
#' @export SparseNeuroVol
#' @import Matrix
#' @details
#' Image data is backed by \code{Matrix::sparseVector}.
#' @examples
#' data <- 1:10
#' indices <- seq(1,1000, length.out=10)
#' bspace <- NeuroSpace(c(64,64,64), spacing=c(1,1,1))
#' sparsevol <- SparseNeuroVol(data,bspace,indices=indices)
#' densevol <- NeuroVol(data,bspace,indices=indices)
#' sum(sparsevol) == sum(densevol)
#'
#'
#'
#' @rdname SparseNeuroVol-class
SparseNeuroVol <- function(data, space, indices=NULL, label="") {
if (length(indices) != length(data)) {
stop(paste("length of 'data' must equal length of 'indices'"))
}
sv <- Matrix::sparseVector(x=data, i=indices, length=prod(dim(space)))
new("SparseNeuroVol", data=sv, space=space)
}
#' LogicalNeuroVol
#'
#' This function constructs a \code{\linkS4class{LogicalNeuroVol}} instance.
#'
#' @param data A three-dimensional \code{array}, a 1D vector with length equal to \code{prod(dim(space))}, or a set of \code{indices} where elements are \code{TRUE}.
#' @param space An instance of class \code{\linkS4class{NeuroSpace}}.
#' @param label A \code{character} string.
#' @param indices An optional 1-d index vector.
#'
#' @return A \code{\linkS4class{LogicalNeuroVol}} instance.
#'
#' @examples
#' # Load an example brain mask
#' brain_mask <- read_vol(system.file("extdata", "global_mask.nii", package="neuroim2"))
#'
#' # Convert the brain mask to a LogicalNeuroVol
#' logical_vol <- LogicalNeuroVol(brain_mask, space(brain_mask))
#'
#' @export LogicalNeuroVol
#' @rdname LogicalNeuroVol-class
LogicalNeuroVol <- function(data, space, label="", indices=NULL) {
if (is.null(dim(data)) && length(data) == prod(dim(space))) {
data <- array(data, dim(space))
} else if (is.null(dim(data)) && !is.null(indices)) {
newdat <- array(FALSE, dim(space))
newdat[indices] <- data
data <- newdat
}
if (length(dim(data)) != 3) {
stop("LogicalNeuroVol: data argument must have three dimensions")
}
if (ndim(space) != 3) {
stop("LogicalVolume: space argument must have three dimensions")
}
if (!is.logical(data)) {
D <- dim(data)
data <- as.logical(data > 0)
dim(data) <- D
}
new("LogicalNeuroVol", .Data=data, space=space)
}
#' Convert DenseNeuroVol to array
#'
#' This function converts a DenseNeuroVol object to an array.
#'
#' @param from A DenseNeuroVol object.
#'
#' @return An array resulting from the conversion.
#'
#' @keywords internal
#' @name as,DenseNeuroVol,array
setAs(from="DenseNeuroVol", to="array", def=function(from) from@.Data)
#' Convert DenseNeuroVol to H5NeuroVol
#'
#' This function converts a DenseNeuroVol object to an H5NeuroVol object.
#'
#' @param from A DenseNeuroVol object.
#'
#' @return An H5NeuroVol object resulting from the conversion.
#'
#' @keywords internal
#' @name DenseNeuroVol,H5NeuroVol
setAs(from="DenseNeuroVol", to="H5NeuroVol", def=function(from) {
to_nih5_vol(from, file_name=NULL, data_type="FLOAT")
})
#' Convert DenseNeuroVec to H5NeuroVec
#'
#' This function converts a DenseNeuroVec object to an H5NeuroVec object.
#'
#' @param from A DenseNeuroVec object.
#'
#' @return An H5NeuroVec object resulting from the conversion.
#'
#' @keywords internal
#' @name DenseNeuroVec,H5NeuroVec
setAs(from="DenseNeuroVec", to="H5NeuroVec", def=function(from) {
to_nih5_vec(from, file_name=NULL, data_type="FLOAT")
})
#' Convert SparseNeuroVol to array
#'
#' This function converts a SparseNeuroVol object to an array object.
#'
#' @param from A SparseNeuroVol object.
#'
#' @return An array object resulting from the conversion.
#'
#' @keywords internal
#' @name SparseNeuroVol,array
setAs(from="SparseNeuroVol", to="array", def=function(from) {
vals <- as.numeric(from@data)
array(vals, dim(from))
})
#' Convert SparseNeuroVol to numeric
#'
#' This function converts a SparseNeuroVol object to a numeric object.
#'
#' @param from A SparseNeuroVol object.
#'
#' @return A numeric object resulting from the conversion.
#' @keywords internal
#' @name SparseNeuroVol,numeric
setAs(from="SparseNeuroVol", to="numeric", def=function(from) {
as.numeric(from@data)
})
#' Convert SparseNeuroVol to numeric
#'
#' @rdname as.numeric-methods
#' @param x the object to convert
#' @export
setMethod(f="as.numeric", signature=signature(x = "SparseNeuroVol"), def=function(x) {
as(x, "numeric")
})
#' conversion from \code{\linkS4class{NeuroVol}} to \code{\linkS4class{LogicalNeuroVol}}
#'
#' @name as
#' @export
setAs(from="NeuroVol", to="LogicalNeuroVol", def=function(from) {
LogicalNeuroVol(as.array(from), space(from))
})
#' conversion from \code{\linkS4class{NeuroVol}} to \code{array}
#'
#' @name as
#' @export
setAs(from="NeuroVol", to="array", def=function(from) from[,,])
#' show a \code{NeuroVol}
#' @param object the object
#' @export
setMethod(f="show", signature=signature("NeuroVol"),
def=function(object) {
sp <- space(object)
cat("NeuroVol\n")
cat(" Type :", class(object), "\n")
cat(" Dimension :", dim(object), "\n")
cat(" Spacing :", paste(paste(signif(sp@spacing[1:(length(sp@spacing)-1)],2), " X ", collapse=" "),
sp@spacing[length(sp@spacing)], "\n"))
cat(" Origin :", paste(paste(signif(sp@origin[1:(length(sp@origin)-1)],2), " X ", collapse=" "),
sp@origin[length(sp@origin)], "\n"))
cat(" Axes :", paste(sp@axes@i@axis, sp@axes@j@axis,sp@axes@k@axis), "\n")
}
)
#' show a \code{SparseNeuroVol}
#' @param object the object
#' @export
setMethod(f="show", signature=signature("SparseNeuroVol"),
def=function(object) {
sp <- space(object)
cat("NeuroVol\n")
cat(" Type :", class(object), "\n")
cat(" Dimension :", dim(object), "\n")
cat(" Spacing :", paste(paste(signif(sp@spacing[1:(length(sp@spacing)-1)],2), " X ", collapse=" "),
sp@spacing[length(sp@spacing)], "\n"))
cat(" Origin :", paste(paste(signif(sp@origin[1:(length(sp@origin)-1)],2), " X ", collapse=" "),
sp@origin[length(sp@origin)], "\n"))
cat(" Axes :", paste(sp@axes@i@axis, sp@axes@j@axis,sp@axes@k@axis), "\n")
cat(" Cardinality :", length(Matrix::which(object@data>0)))
}
)
#' load a NeuroVol
#' @noRd
## TODO reduce code duplication with load_data#NeuroVecSource
setMethod(f="load_data", signature=c(x="NeuroVolSource"),
def=function(x) {
meta <- x@meta_info
nels <- prod(meta@dims[1:3])
### for brain buckets, this offset needs to be precomputed ....
offset <- (nels * (x@index-1)) * meta@bytes_per_element
reader <- data_reader(meta, offset)
dat <- read_elements(reader, nels)
## bit of a hack to deal with scale factors
if (.hasSlot(meta, "slope")) {
if (meta@slope != 0) {
dat <- dat*meta@slope
}
}
close(reader)
arr <- array(dat, meta@dims[1:3])
bspace <- NeuroSpace(meta@dims[1:3], meta@spacing, meta@origin, meta@spatial_axes, trans(meta))
DenseNeuroVol(arr, bspace, x)
})
#' Constructor for NeuroVolSource
#'
#' @param input the input file name
#' @param index the image subvolume index
#' @export
#' @rdname NeuroVolSource-class
#' @return a new instance of type \code{NeuroVolSource}
NeuroVolSource <- function(input, index=1) {
stopifnot(index >= 1)
stopifnot(is.character(input))
if (!file.exists(input)) {
candidates <- Sys.glob(paste(input, "*", sep=""))
if (length(candidates) > 0) {
input <- candidates[1]
}
}
stopifnot(file.exists(input))
meta_info <- read_header(input)
if ( length(meta_info@dims) < 4 && index > 1) {
stop("index cannot be greater than 1 for a image with dimensions < 4")
}
new("NeuroVolSource", meta_info=meta_info, index=as.integer(index))
}
#' Load an image volume from a file
#'
#' @param file_name the name of the file to load
#' @param index the index of the volume (e.g. if the file is 4-dimensional)
#' @return an instance of the class \code{\linkS4class{DenseNeuroVol}}
#'
#' @examples
#' fname <- system.file("extdata", "global_mask.nii", package="neuroim2")
#' x <- read_vol(fname)
#' print(dim(x))
#' space(x)
#'
#' @export read_vol
read_vol <- function(file_name, index=1) {
src <- NeuroVolSource(file_name, index)
load_data(src)
}
#' @export
#' @rdname slices-methods
setMethod(f="slices", signature=signature(x="NeuroVol"),
def = function(x) {
nslices <- dim(x)[3]
f <- function(i) slice(x, i, 3)
#lis <- lapply(1:nslices, function(i) f)
#deferred_list(lis)
deflist::deflist(f, nslices)
})
#' @export
#' @rdname extractor3d
setMethod(f="[", signature=signature(x = "NeuroVol", i = "ROICoords", j = "missing"),
def=function (x, i, j, k, ..., drop=TRUE) {
callGeneric(x, i@coords)
}
)
#' @rdname concat-methods
#' @export
setMethod(f="concat", signature=signature(x="DenseNeuroVol", y="missing"),
def=function(x,y,...) {
x
})
#' @note dimensions of x and y must be equal
#' @export
#' @rdname concat-methods
setMethod(f="concat", signature=signature(x="DenseNeuroVol", y="DenseNeuroVol"),
def=function(x,y,...) {
.concat4D(x,y,...)
})
#' @export
#' @rdname map_values-methods
setMethod(f="map_values", signature=signature(x="NeuroVol", lookup="list"),
def=function(x,lookup) {
if (is.null(names(lookup))) {
names(lookup) <- seq_along(lookup)
}
out <- match(x,as.numeric(names(lookup)))
DenseNeuroVol(unlist(lookup[out]), space(x))
})
#' @export
#' @rdname map_values-methods
setMethod(f="map_values", signature=signature(x="NeuroVol", lookup="matrix"),
def=function(x,lookup) {
if (ncol(lookup) != 2) {
stop("map_values: lookup matrix must have two columns: column 1 is key, column 2 is value")
} else if (nrow(lookup) < 1) {
stop("map_values: lookup matrix have at least one row")
}
m <- match(as.vector(x), lookup[,1])
outv <- lookup[m,2]
outv[is.na(outv)] <- 0
NeuroVol(array(outv, dim(x)), space(x))
})
#' @export
#' @rdname split_fill-methods
setMethod(f="split_fill", signature=signature(x="NeuroVol", fac="factor", FUN="function"),
def=function(x,fac,FUN) {
stopifnot(length(x) == length(fac))
S <- split(1:length(x), fac, drop=TRUE)
res <- map(S, function(ind) {
X <- FUN(x[ind])
if (length(X) == length(ind)) {
X
} else {
rep(X, length(ind))
}
})
ovol <- x
ovol[1:length(x)] <- unsplit(res, fac)
ovol
})
#' @keywords internal
#' @noRd
fast.expand.grid <- function(seq1,seq2, constant) {
cbind(Var1 = rep.int(seq1, length(seq2)),
Var2 = rep.int(seq2, rep.int(length(seq1),length(seq2))),
Var3=constant)
}
#' @export
#' @rdname slice-methods
setMethod(f="slice", signature=signature(x="NeuroVol", zlevel="numeric", along="numeric",
orientation="missing"),
def=function(x, zlevel, along, orientation) {
stopifnot(along >= 1 && along <=3)
stopifnot(zlevel >= 1 && zlevel <= dim(x)[along])
imslice <- switch(as.character(along),
"1"=x[zlevel,,],
"2"=x[,zlevel,],
"3"=x[,,zlevel])
NeuroSlice(imslice, drop_dim(space(x), along))
})
#' @export
#' @rdname slice-methods
setMethod(f="slice", signature=signature(x="NeuroVol", zlevel="numeric", along="NeuroSpace",
orientation="AxisSet3D"),
def=function(x, zlevel, along, orientation) {
xdim <- dim_of(along, orientation@i)
ydim <- dim_of(along, orientation@j)
zdim <- dim_of(along, orientation@k)
#browser()
stopifnot(zlevel >= 1 && zlevel <= zdim)
vox <- as.matrix(fast.expand.grid(seq(1,xdim), seq(1,ydim), zlevel))
gg <- grid_to_grid(along, vox)
imslice <- x[gg]
NeuroSlice(matrix(imslice, xdim,ydim), drop_dim(along,which_dim(along, orientation@k)))
})
#' @title Convert NeuroVol to a mask
#' @description This method converts a NeuroVol object to a mask by setting all positive values to TRUE and all non-positive values to FALSE.
#' @param x A NeuroVol object to convert to a mask.
#' @param indices A missing argument; not used in this method.
#' @return A LogicalNeuroVol object representing the mask created from the input NeuroVol.
#' @export
#' @rdname as.mask-methods
setMethod(f="as.mask", signature=signature(x="NeuroVol", indices="missing"),
def=function(x) {
LogicalNeuroVol(x > 0, space(x))
})
#' @title Convert NeuroVol to a mask with specified indices
#' @description This method converts a NeuroVol object to a mask by setting the specified indices to TRUE and the remaining elements to FALSE.
#' @param x A NeuroVol object to convert to a mask.
#' @param indices A numeric vector containing the indices of the input NeuroVol that should be set to TRUE in the resulting mask.
#' @return A LogicalNeuroVol object representing the mask created from the input NeuroVol with specified indices.
#' @export
#' @rdname as.mask-methods
setMethod(f="as.mask", signature=signature(x="NeuroVol", indices="numeric"),
def=function(x, indices) {
M <- array(0, dim(x))
M[indices] <- 1
LogicalNeuroVol(M, space(x))
})
#' @export
#' @rdname coord_to_grid-methods
setMethod(f="coord_to_grid", signature=signature(x="NeuroVol", coords="matrix"),
def=function(x, coords) {
callGeneric(space(x), coords)
})
#' @export
#' @rdname coord_to_grid-methods
setMethod(f="coord_to_grid", signature=signature(x="NeuroVol", coords="numeric"),
def=function(x, coords) {
coords <- matrix(coords, nrow=1)
callGeneric(x, coords)
})
#' @importFrom dbscan kNN
#' @keywords internal
#' @noRd
.pruneCoords <- function(coord.set, vals, mindist=10) {
if (NROW(coord.set) == 1) {
1
}
.prune <- function(keepIndices) {
if (length(keepIndices) <= 2) {
keepIndices
} else {
ret <- try(dbscan::kNN(coord.set[keepIndices,], coord.set[keepIndices,], k=2))
#ret <- rflann::Neighbour(coord.set[keepIndices,], coord.set[keepIndices,], k=2, build="kdtree", cores=0, checks=1)
#ind <- ret$indices[, 2]
#ds <- sqrt(ret$distances[, 2])
ind <- ret$id[,2]
ds <- ret$distances[,2]
v <- vals[keepIndices]
ovals <- v[ind]
pruneSet <- ifelse(ds < mindist & ovals > v, TRUE, FALSE)
if (any(pruneSet)) {
Recall(keepIndices[!pruneSet])
} else {
keepIndices
}
}
}
.prune(1:NROW(coord.set))
}
#' @title Create a patch set from a NeuroVol object
#' @description This function creates a patch set from a NeuroVol object given specified dimensions
#' @return A deferred list of patches.
#' @param x a NeuroVol object
#' @param dims the dimensions of the patch
#' @param ... additional args
#' @export
#' @rdname patch_set-methods
setMethod(f="patch_set", signature=signature(x="NeuroVol",
dims="numeric",
mask="missing"),
def=function(x, dims, ...) {
mask <- LogicalNeuroVol(array(1, dim(x)), space=space(x))
callGeneric(x, dims, mask)
})
#' @title Create a patch set from a NeuroVol object
#' @description This function creates a patch set from a NeuroVol object given specified dimensions and a mask.
#' @return A deferred list of patches.
#' @param x a NeuroVol object
#' @param dims the dimensions of the patch
#' @param mask the mask defining the valid patch centers
#' @param ... additional args
#' @export
#' @rdname patch_set-methods
setMethod(f="patch_set", signature=signature(x="NeuroVol",
dims="numeric",
mask="LogicalNeuroVol"),
def=function(x, dims, mask, ...) {
if (!all(dims > 0)) {
stop("all 'dims' must be greater than zero")
}
if (!all(dims %% 2 == 1)) {
stop("all 'dims' must be odd numbers")
}
template <- as.matrix(expand.grid
(x=seq(seq(ceiling(-dims[1]/2),floor(dims[1]/2))),
y=seq(seq(ceiling(-dims[2]/2),floor(dims[2]/2))),
z=seq(seq(ceiling(-dims[3]/2),floor(dims[3]/2)))))
xdim <- dim(x)
grid <- index_to_grid(mask, which(mask>0))
f <- function(i) {
g <- grid[i,]
m <- t(t(template) + g)
m[,1] <- pmax(pmin(m[,1], xdim[1]), 1)
m[,2] <- pmax(pmin(m[,2], xdim[2]), 1)
m[,3] <- pmax(pmin(m[,3], xdim[3]), 1)
ret <- x[m]
attr(ret, "idx") <- grid_to_index(x, m)
ret
}
patches <- deflist::deflist(f, nrow(grid))
patches
#patches <- deferred_list(lapply(1:nrow(grid), function(i) f))
})
#' apply a kernel function to a \code{\linkS4class{NeuroVol}}
#'
#' @rdname map-methods
#' @param mask restrict application of kernel to masked area
#' @export
setMethod(f="mapf", signature=signature(x="NeuroVol", m="Kernel"),
def=function(x, m, mask=NULL) {
ovol <- array(0, dim(x))
hwidth <- map_dbl(m@width, function(d) ceiling(d/2 -1)) + 1
xdim <- dim(x)[1]
ydim <- dim(x)[2]
zdim <- dim(x)[3]
if (!is.null(mask)) {
if (!all.equal(dim(mask), dim(ovol))) {
stop(paste("mask must have same dimensions as input volume"))
}
grid <- index_to_grid(mask, which(mask != 0))
} else {
grid <- as.matrix(expand.grid(i=hwidth[1]:(xdim - hwidth[1]),
j=hwidth[2]:(ydim - hwidth[2]),
k=hwidth[3]:(zdim - hwidth[3])))
}
res <- apply(grid, 1, function(vox) {
loc <- t(t(m@voxels) +vox)
ivals <- x[loc]
sum(ivals * m@weights)
})
ovol[grid] <- res
NeuroVol(ovol, space(x))
})
#' find connected components in NeuroVol
#'
#' @export
#' @importFrom purrr map_int
#' @param threshold threshold defining lower intensity bound for image mask
#' @param cluster_table return cluster_table
#' @param local_maxima return table of local maxima
#' @param local_maxima_dist the distance used to define minum distance between local maxima
#' @rdname conn_comp-methods
setMethod(f="conn_comp", signature=signature(x="NeuroVol"),
def=function(x, threshold=0, cluster_table=TRUE, local_maxima=TRUE, local_maxima_dist=15,...) {
mask <- (x > threshold)
stopifnot(any(mask))
comps <- conn_comp_3D(mask,...)
grid <- as.data.frame(index_to_grid(mask, which(mask>0)))
colnames(grid) <- c("x", "y", "z")
locations <- split(grid, comps$index[comps$index>0])
ret <- list(index=ClusteredNeuroVol(mask, clusters=comps$index[mask>0]), size=NeuroVol(comps$size, space(x)), voxels=locations)
if (cluster_table) {
maxima <- do.call(rbind, lapply(locations, function(loc) {
if (nrow(loc) == 1) {
loc
} else {
vals <- x[as.matrix(loc)]
loc[which.max(vals),]
}
}))
N <- comps$size[as.matrix(maxima)]
Area <- N * prod(spacing(x))
maxvals <- x[as.matrix(maxima)]
ret$cluster_table <- data.frame(index=1:NROW(maxima),
x=maxima[,1], y=maxima[,2], z=maxima[,3],
N=N, Area=Area, value=maxvals)
}
if (local_maxima) {
#if (all(map_int(locations, NROW) == 1)) {
#
#}
coord.sets <- lapply(locations, function(loc) {
sweep(as.matrix(loc), 2, spacing(x), "*")
})
loc.max <- do.call(rbind, mapply(function(cset, i) {
idx <- .pruneCoords(as.matrix(cset), x[as.matrix(locations[[i]])], mindist=local_maxima_dist)
maxvox <- as.matrix(locations[[i]])[idx,,drop=F]
cbind(i, maxvox)
}, coord.sets, 1:length(coord.sets), SIMPLIFY=FALSE))
loc.max <- cbind(loc.max, x[loc.max[, 2:4, drop=F]])
row.names(loc.max) <- 1:NROW(loc.max)
colnames(loc.max) <- c("index", "x", "y", "z", "value")
ret$local_maxima <- loc.max
}
ret
})
### TODO when the source voulme is an AFNI BRIK the output file does not preserve orientation
### e.g. write_vol(x, space(BRIK), ...)
### if BRIK was RAI, output NIFTI file reverts to LPI
#' @export
#' @rdname write_vol-methods
setMethod(f="write_vol",signature=signature(x="NeuroVol", file_name="character", format="missing", data_type="missing"),
def=function(x, file_name) {
write_nifti_volume(x, file_name)
})
#' @export
#' @rdname write_vol-methods
setMethod(f="write_vol",signature=signature(x="ClusteredNeuroVol", file_name="character", format="missing", data_type="missing"),
def=function(x, file_name) {
callGeneric(as(x, "DenseNeuroVol"), file_name)
})
#' @export
#' @rdname write_vol-methods
setMethod(f="write_vol",signature=signature(x="NeuroVol", file_name="character", format="character", data_type="missing"),
def=function(x, file_name, format) {
if (toupper(format) == "NIFTI" || toupper(format) == "NIFTI1" || toupper(format) == "NIFTI-1") {
callGeneric(x, file_name)
} else {
stop(paste("sorry, cannot write format: ", format))
}
})
#' @export
#' @rdname write_vol-methods
setMethod(f="write_vol",signature=signature(x="ROIVol", file_name="character", format="character", data_type="missing"),
def=function(x, file_name, format) {
if (toupper(format) == "NIFTI" || toupper(format) == "NIFTI1" || toupper(format) == "NIFTI-1") {
callGeneric(as.dense(x), file_name)
} else {
stop(paste("sorry, cannot write format: ", format))
}
})
#' @export write_vol
#' @rdname write_vol-methods
setMethod(f="write_vol",signature=signature(x="NeuroVol", file_name="character", format="missing", data_type="character"),
def=function(x, file_name, data_type) {
write_nifti_volume(x, file_name, data_type)
})
#' as.logical
#'
#' Convert NeuroVol to \code{\linkS4class{LogicalNeuroVol}}
#'
#' the image values will be converted to using R base function \code{as.logical} and wrapped in \code{LogicalNeuroVol}
#'
#' @param x the object
#' @return an instance of \code{\linkS4class{LogicalNeuroVol}}
#' @rdname as.logical-methods
#' @export
setMethod(f="as.logical", signature=signature(x = "NeuroVol"), def=function(x) {
vals <- as.logical(as.vector(x))
LogicalNeuroVol(vals, space(x))
})
#' @rdname as.sparse-methods
setMethod(f="as.sparse", signature=signature(x="DenseNeuroVol", mask="LogicalNeuroVol"),
def=function(x, mask) {
assert_that(all(dim(x) == dim(mask)))
assert_that(all(spacing(x) == spacing(mask)))
dat <- x[mask]
bvec <- SparseNeuroVol(data=dat, space=space(x),indices=which(mask>0))
})
#' @rdname partition-methods
setMethod("partition", signature=signature(x="LogicalNeuroVol", k="integer"),
def=function(x,k) {
assert_that(k>1)
idx <- which(x != 0)
cds <- index_to_coord(x,idx)
kres <- kmeans(cds, centers=k, iter.max=200)
ClusteredNeuroVol(x, kres$cluster)
})
#' @rdname partition-methods
setMethod("partition", signature=signature(x="LogicalNeuroVol", k="numeric"),
def=function(x,k) {
callGeneric(x, as.integer(k))
})
#' @rdname partition-methods
setMethod("partition", signature=signature(x="DenseNeuroVol", k="numeric"),
def=function(x,k) {
callGeneric(as.logical(x), as.integer(k))
})
#' @rdname as.sparse-methods
setMethod(f="as.sparse", signature=signature(x="DenseNeuroVol", mask="numeric"),
def=function(x, mask) {
m <- as.integer(mask)
bvol <- SparseNeuroVol(x[m], space(x), indices=m)
bvol
})
#' @noRd
setMethod(f="linear_access", signature=signature(x = "SparseNeuroVol", i = "numeric"),
def=function (x, i) {
x@data[as.numeric(i)]
})
#' extractor
#' @export
#' @param x the object
#' @param i first index
#' @param j second index
#' @param k third index
#' @param ... additional args
#' @param drop dimension
setMethod(f="[", signature=signature(x = "SparseNeuroVol", i = "numeric", j = "numeric", drop="ANY"),
def=function (x, i, j, k, ..., drop=TRUE) {
if (missing(k)) {
k <- seq(1, dim(x)[3])
}
nel <- length(i) * length(j) * length(k)
mind <- cbind(rep(i, length.out=nel), rep(j, each=length(i)), rep(k, each=length(i) * length(j)))
ind <- grid_to_index(x, mind)
dat <- x@data[as.numeric(ind)]
array(dat, c(length(i), length(j), length(k)))
})
#' plot an NeuroVol as a series of 2D slices
#'
#' @rdname plot-methods
#' @param x the object to display
#' @param cmap a color map consisting of a vector of colors in hex format (e.g. \code{gray(n=255)})
#' @param zlevels the series of slice indices to display.
#' @param irange the intensity range indicating the low and high values of the color scale.
#' @param thresh a 2-element vector indicating the lower and upper transparency thresholds.
#' @param alpha the level of alpha transparency
#' @param bgvol a background volume that serves as an image underlay (currently ignored).
#' @param bgcmap a color map for backround layer consisting of a vector of colors in hex format (e.g. \code{gray(n=255)})
#' @export
#' @importFrom graphics plot
#' @import colorplane
#' @examples
#'
#' dat <- matrix(rnorm(100*100), 100, 100)
#' slice <- NeuroSlice(dat, NeuroSpace(c(100,100)))
#' #plot(slice)
setMethod("plot", signature=signature(x="NeuroVol"),
def=function(x,cmap=gray(seq(0,1,length.out=255)),
zlevels=unique(round(seq(1, dim(x)[3], length.out=6))),
irange=range(x, na.rm=TRUE),
thresh=c(0,0),
alpha=1,
bgvol=NULL,
bgcmap=gray(seq(0,1,length.out=255))) {
if (!requireNamespace("ggplot2", quietly = TRUE)) {
stop("Package \"ggplot2\" needed for this function to work. Please install it.",
call. = FALSE)
}
if (!is.null(bgvol)) {
assert_that(all(dim(x) == dim(bgvol)))
assert_that(all(spacing(x) == spacing(bgvol)))
}
df1 <- do.call(rbind, purrr::map(zlevels, function(i) {
if (!is.null(bgvol)) {
bgslice <- slice(bgvol, zlevel=i, along=3)
bgplane <- colorplane::IntensityColorPlane(as.numeric(bgslice), cols=bgcmap)
bgcols <- colorplane::map_colors(bgplane)
}
imslice <- slice(x, zlevel=i, along=3)
implane <- colorplane::IntensityColorPlane(as.numeric(imslice), cols=cmap, alpha=alpha)
fgcols <- colorplane::map_colors(implane, threshold=thresh, irange=irange)
if (!is.null(bgvol)) {
fgcols <- colorplane::as_hexcol(colorplane::blend_colors(bgcols, fgcols, alpha=alpha))
} else {
fgcols <- colorplane::as_hexcol(fgcols)
}
cds <- index_to_coord(space(imslice), 1:length(imslice))
data.frame(x=cds[,1], y=cds[,2], z=i, value=as.vector(fgcols))
}))
{y=value=NULL}
p <-
ggplot2::ggplot(ggplot2::aes(x=x, y=y), data=df1) +
ggplot2::coord_fixed() +
ggplot2::geom_raster(ggplot2::aes(x=x,y=y, fill=value)) +
ggplot2::scale_fill_identity() + ggplot2::xlab("") + ggplot2::ylab("") +
ggplot2::scale_x_continuous(expand=c(0,0)) +
ggplot2::scale_y_continuous(expand=c(0,0)) +
ggplot2::facet_wrap(~ z) +
ggplot2::theme_classic()
p
})
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.