Package: segmantR
Title: Cell Segmentation for Histology with Human-in-the-Loop Training
Version: 0.2.0
Authors@R: 
    person("R.", "Heller", , "raban.heller@uni-ulm.de",
           role = c("aut", "cre", "cph"),
           comment = c(ORCID = "0000-0001-8006-9742"))
Description: Cell segmentation for histology and multiplexed tissue
        imaging. Provides classical segmentation methods (adaptive
        thresholding, watershed, Voronoi propagation), wrappers through
        'reticulate' for the deep learning models 'Cellpose' (Stringer
        et al. (2021) <doi:10.1038/s41592-020-01018-x>), 'StarDist'
        (Schmidt et al. (2018) <doi:10.1007/978-3-030-00934-2_30>) and
        'Mesmer' (Greenwald et al. (2022)
        <doi:10.1038/s41587-021-01094-0>), and a human-in-the-loop
        iterative refinement workflow. A 'shiny' application lets
        researchers annotate, correct, train, evaluate and download
        custom segmentation models without leaving R. Versioned
        declarative segmentation protocols, a neutral interchange
        format (JSON manifests, integer label TIFF, GeoJSON,
        measurement tables with SHA-256 inventories), training dataset
        manifests with grouped splits, and readable 'QuPath' script
        templates with a 'StarDist' parameter adapter support
        reproducible exchange with 'QuPath' and partner packages.
License: MIT + file LICENSE
URL: https://github.com/CTTIR/segmantR,
        https://cttir.github.io/segmantR/
BugReports: https://github.com/CTTIR/segmantR/issues
Encoding: UTF-8
Language: en-GB
SystemRequirements: Optional: Python with the 'cellpose', 'stardist' or
        'deepcell' modules for the deep learning backends; QuPath (>=
        0.4) to run the exported QuPath scripts.
Roxygen: list(markdown = TRUE)
Depends: R (>= 4.1.0)
Imports: digest, dplyr, rlang, sf (>= 1.0.0), stars (>= 0.6.0), cli,
        ggplot2 (>= 3.4.0), graphics, grDevices, jsonlite, shiny (>=
        1.7.0), stats, tibble, tools, utils
Suggests: callr, EBImage, htmlwidgets, imager, knitr, magick,
        nanoparquet, openssl, processx, reticulate (>= 1.35),
        rmarkdown, terra, testthat (>= 3.0.0), tiff, withr, zip
Config/testthat/edition: 3
VignetteBuilder: knitr
Config/Needs/website: CTTIR/themakR
Config/roxygen2/version: 8.0.0
Config/pak/sysreqs: libabsl-dev cmake libgdal-dev gdal-bin libgeos-dev
        make libuv1-dev libssl-dev libproj-dev python3 libsqlite3-dev
        libudunits2-dev zlib1g-dev
Repository: https://cttir.r-universe.dev
Date/Publication: 2026-10-29 11:19:01 UTC
RemoteUrl: https://github.com/CTTIR/segmantR
RemoteRef: HEAD
RemoteSha: 882c2fb7fe0546f71f6f11bbfd3de6103859b098
NeedsCompilation: no
Packaged: 2026-11-03 18:19:54 UTC; root
Author: R. Heller [aut, cre, cph] (ORCID:
    <https://orcid.org/0000-0001-8006-9742>)
Maintainer: R. Heller <raban.heller@uni-ulm.de>