Journal: Frontiers in computer science

Article Title: Integration of the ImageJ Ecosystem in the KNIME Analytics Platform

doi: 10.3389/fcomp.2020.00008

Figure Lengend Snippet: Side-by-side comparison of ImageJ macro with KNIME workflow using KNIME Image Processing nodes.

Article Snippet: The KNIME Image Processing extension is built on the same technologies that have driven the latest versions of ImageJ, including ImgLib2, a general-purpose, multidimensional image processing library ( Pietzsch et al. 2012 ); SCIFIO, a framework for reading and writing scientific image formats ( Hiner, Rueden, and Eliceiri 2016 ); Bio-Formats, a Java library for reading and writing life sciences image file formats ( Linkert et al. 2010 ); and BigDataViewer ( Pietzsch et al. 2015 ), a re-slicing browser for terabyte-sized multi-view image sequences.

Techniques: Comparison

Journal: Frontiers in computer science

Article Title: Integration of the ImageJ Ecosystem in the KNIME Analytics Platform

doi: 10.3389/fcomp.2020.00008

Figure Lengend Snippet: Quantitative analysis of subcellular structures. To analyze the consequences of disrupting the cytoskeleton on matrix adhesion focal adhesion complex formation is visualized using TIRF imaging. A KNIME workflow was leveraged to define the number of focal adhesion complexes at the periphery of the cell vs. within the center of the cell using sequential steps with discrete objectives: I. Read file in, II. Processing of the images into labels that capture the individual focal adhesion complexes, III. Arithmetic on the labels to obtain the characteristics (features) for classification, IV. Classification of the labels using extracted features, V. Visualization of the classified focal adhesion complexes (periphery vs center). Note that the output of this pipeline is a combination of visualization and quantitation. Both of these can be leveraged for further analysis.

Article Snippet: The KNIME Image Processing extension is built on the same technologies that have driven the latest versions of ImageJ, including ImgLib2, a general-purpose, multidimensional image processing library ( Pietzsch et al. 2012 ); SCIFIO, a framework for reading and writing scientific image formats ( Hiner, Rueden, and Eliceiri 2016 ); Bio-Formats, a Java library for reading and writing life sciences image file formats ( Linkert et al. 2010 ); and BigDataViewer ( Pietzsch et al. 2015 ), a re-slicing browser for terabyte-sized multi-view image sequences.

Techniques: Imaging, Quantitation Assay

Journal: Frontiers in computer science

Article Title: Integration of the ImageJ Ecosystem in the KNIME Analytics Platform

doi: 10.3389/fcomp.2020.00008

Figure Lengend Snippet: Quantitative analysis of histological stain. The histological staining of a cell adhesion marker (CD166) related to tumor invasion and metastasis demonstrates significant variation across patient samples. As in user case #1, the KNIME workflow was divided into sequential steps that complete discrete objectives: 1) read in file , 2) pre-processing of the images and their annotation in preparation for analysis using ImageJ2 functionalities, 3) pixel classification using Weka-bases machine learning functionality, 4) post-classification processing of image data to labels that correspond to ‘positive’, 5) compilation of labels, images and annotations, 6) visualization of the quantitation by overlaying the labels with the original image.

Article Snippet: The KNIME Image Processing extension is built on the same technologies that have driven the latest versions of ImageJ, including ImgLib2, a general-purpose, multidimensional image processing library ( Pietzsch et al. 2012 ); SCIFIO, a framework for reading and writing scientific image formats ( Hiner, Rueden, and Eliceiri 2016 ); Bio-Formats, a Java library for reading and writing life sciences image file formats ( Linkert et al. 2010 ); and BigDataViewer ( Pietzsch et al. 2015 ), a re-slicing browser for terabyte-sized multi-view image sequences.

Techniques: Staining, Marker, Quantitation Assay