Getting Started with hyperspectR

R-CMD-check pkgdown CRAN status Codecov test coverage CRAN downloads CRAN downloads total License: MIT Lifecycle: experimental

Installation

# Install from GitHub
# install.packages("remotes")
remotes::install_github("cttir/hyperspectR")

Loading the Package

library(hyperspectR)

Creating and Inspecting a Cube

hyperspectR provides synthetic data generators for testing and demonstration. hs_example_cube() creates a 30x30 pixel tissue scene with 61 spectral bands spanning 430-910 nm:

cube <- hs_example_cube()
print(cube)
#> 
#> ── hsi_cube ────────────────────────────────────────────────────────────────────
#> Dimensions: 30 rows x 30 cols x 61 bands
#> Wavelengths: 430-910 nm (61 bands)
#> FWHM: 25 nm (mean)
#> Mask: 900/900 valid pixels (100%)
#> Data range: [0.0198, 0.7306]
#> Metadata: camera, spectral_model, interpretation, processing_mode,
#> acquisition_time, region_map, sto2_ground_truth, seed

Check the dimensions:

dim(cube)
#> [1] 30 30 61

The summary provides per-band statistics:

s <- summary(cube)
s$wavelength_range
#> [1] 430 910
s$data_range
#> [1] 0.01982067 0.73056093

Subsetting

You can subset cubes spatially and spectrally while preserving the class:

sub <- cube[1:15, 1:15, 1:30]
dim(sub)
#> [1] 15 15 30

Visualization

RGB Composite

ggplot2::autoplot(cube, type = "rgb")

Single Band Image

hs_plot_image(cube, wavelength = 550)

Mean Spectrum

hs_plot_spectra(cube, pixels = "mean", show_sd = TRUE)

Computing Tissue Indices

The package provides a research hemoglobin fraction estimate and dimensionless band ratios. These outputs are not validated clinical measurements:

sto2 <- hs_sto2(cube)
hs_plot_index(sto2, title = "Fitted Hb fraction (%)", palette = "sto2")

Clinical Panel Display

hs_plot_clinical(cube, indices = c("sto2", "npi", "thi"))

Interactive Exploration

Launch the Shiny app for interactive analysis:

hs_run_app(cube)

Next Steps

  • Read real data with hs_read_envi() or hs_read_cube()
  • Preprocess with hs_smooth(), hs_snv(), hs_msc()
  • Perform PCA with hs_pca() or MNF with hs_mnf()
  • Classify with hs_sam() or fit chromophores with hs_beer_lambert()

Citation

To cite hyperspectR in publications, please use:

@Manual{,
  title = {hyperspectR: Hyperspectral Imaging Analysis for Biomedical Applications},
  author = {R. Heller and V. Forstmeier},
  year = {2026},
  note = {R package version 0.1.0},
  url = {https://github.com/cttir/hyperspectR},
}