biopixR: Extracting Insights from BiologicalImages

biopixR is an R package designed specifically for the analysis of bioimage data.The package comprises 15 functions that cater to a variety of tasks, including image import, segmentation, feature extraction, quantification, clustering, parameter optimization, and a distinctive algorithm for line gap mitigation.Among these is an optimized edge detection algorithm based on the Canny edge detection algorithm.Together with a noise-resilient threshold method, these two algorithms form the foundation of feature extraction within the biopixR package.Both edge detection and thresholding can be parameterized via an interactive Tcl/Tk user interface.For circular objects, the threshold adjustment factor and the smoothing factor for edge detection can be automatically calculated using Gaussian process regression models (Kriging models).Thereby making an automated feature extraction, which includes segmentation and labeling, available in R. Furthermore, all available methods for feature extraction and filtering can be accessed within a single pipeline function, not found in other packages.The package also includes a function for clustering objects within an image based on their shape features using Self-Organizing Maps (SOM).Designed for medium-throughput analysis, the package is able to analyze whole directories, utilize multiple cores for parallel processing, and generate detailed log files to track the analytical process.Furthermore, the biopixR package offers distinctive data sets of microbead images and microbeads in water-oil emulsions.The fundamental functionality of biopixR was recently employed in the study by Geithe et al. (2024) to perform quality control on microbeads.As the biopixR package provides access to a range of fundamental data, including object size, quantity, shape, and intensity, it is a valuable tool for researchers in the field.The biopixR package leverages capabilities from the imager (Barthelm & Tschumperl, 2019) and magick (Ooms, 2024) packages to perform its own tasks related to biological image processing and analysis.