Qiber3D—an open-source software package for the quantitative analysis of networks from 3D image stacks

Abstract

Background Optical slice microscopy is commonly used to observe cellular morphology in 3D tissue culture, e.g., the formation of cell-derived networks. The morphometric quantification of these networks is essential to study the cellular phenotype. Commonly, the quantitative measurements are performed on 2D projections of the image stack, resulting in the loss of information in the third dimension. Currently available 3D image analysis tools rely on manual interactions with the software and are therefore not feasible for large datasets. Findings Here we present Qiber3D, an open-source image processing toolkit. The software package includes the essential image analysis procedures required for image processing, from the raw image to the quantified data. Optional pre-processing steps can be switched on/off depending on the input data to allow for analyzing networks from a variety of sources. Two reconstruction algorithms are offered to meet the requirements for a wide range of network types. Furthermore, Qiber3D’s rendering capabilities enable the user to inspect each step of the image analysis process interactively to ensure the creation of an optimal workflow for each application. Conclusions Qiber3D is implemented as a Python package, and its source code is freely available at https://github.com/theia-dev/Qiber3D. The toolkit was designed using a building block principle to enable the analysis of a variety of structures, such as vascular networks, neuronal structures, or scaffolds from numerous input formats. While Qiber3D can be used interactively in the Python console, it is aimed at unsupervised automation to process large image datasets efficiently.

Recovery of Therapeutically Ablated Engineered Blood‐Vessel Networks on a Plug‐and‐Play Platform

“…This allows for the fast acquisition of images at ambient conditions during culture without overt loss in viability and reduces required data storage capacity. However, 3D quantification, for instance using the recently published Qiber3D [ 50 ] method is compatible with our model. Not surprisingly, comparing 3D and 2D quantification of networks after 2 days of ablation or control treatment, it was evident that the total length is underestimated in 2D quantifications.…”

“…Interestingly, when the lower dose of 10 µg mL −1 bevacizumab was applied, the network length did not appear to decrease. In addition, on day 6 of co‐culture, the total network length of control and high‐dose samples was quantified both from matched maximum intensity projections (“2D”) using Angiogenesis Analyzer for ImageJ [ 49 ] and full z‐stacks (“3D”) based on a method described elsewhere [ 50 ] (Figure 3f). Interestingly, while the 2D quantification accounted for 65% of the total 3D length for control samples and 76% for ablated samples, a significant decrease in total structure length was evident with both methods.…”

“…This region was applied to maximum projections of all channels of interest and the intensity distribution was assessed using a “plot profile.” The 3D GFP‐HUVEC network quantification was done using the Qiber3D Python package according to Jaeschke et al. [ 50 ] The default settings were adapted as follows: a median filter was applied in the pre‐processing, but no intensity attenuation correction or Gaussian filter was applied. Binarization was done using the OTSU algorithm.…”

A systematic review and comparison of automated tools for quantification of fibrous networks

“…FIRE 23 was not used, since CT-FIRE 24 was used instead, which is an extension of FIRE. Qiber3D 25 and the 3D directional variance algorithm 26 could only be used on 3D image stacks. However, all other tools were applied to 2D images, so we did not include Qiber3D and the 3D directional variance algorithm in the comparison analyses.…”

Beyond Your Sight Using Metaverse Immersive Vision With Technology Behaviour Model

“…The use of multiinteraction technologies, especially vision and haptics, helps the surgical expert to operate with greater precision. HVEU is mainly based on digital to visual 360 image display in surgery (Jaeschke et al, 2022), combined with medical images to achieve actual tissue images of different patients. It is suitable for remote real-world transmission, voice communication, visual manipulation, and haptic sensing, facilitating better than traditional surgical operations (Cofano et al, 2021;Kwok & Koh, 2020).…”