Reads QuPath 0.4+ exports (a FeatureCollection or a bare feature array) and
the legacy annotatR 0.1 dialect. classification.name (or the joined
names of a derived class, e.g. "Tumor: Positive") becomes the ROI label,
the classification colour the layer style colour, and features without a
classification are labelled "unclassified". Everything else QuPath carries
is kept, never invented, under each ROI's attributes$qupath: the object
id and type, name, measurements (with explicit value_state), the image
plane (zero-based c, z, t; c = NA for all channels), metadata, a
nucleus geometry, and roi_native = "ellipse" with
geometry_fidelity = "approximated" for polygonised ellipses. An segmantR id
stored in metadata.annotatr_roi_id is restored as the ROI id.
Usage
sg_roi_read_qupath(
path,
layer_name = "qupath",
level = 0L,
call = rlang::caller_env()
)Value
An sg_roi_layer.
Examples
lyr <- sg_roi_layer_add(sg_roi_layer("tissue"),
sg_roi_rect(2, 2, 8, 8, label = "tissue"))
p <- tempfile(fileext = ".geojson")
sg_roi_write_qupath(sg_roi_project(sg_roi_frame(c(20, 20)), lyr), p)
sg_roi_layer_n(sg_roi_read_qupath(p))
#> [1] 1