File Menu

Provides access to some file handling actions for image dataset

Back to Index --> User Guide --> Menu

collapse all on page

Contents

Import image from...

collapse all

Example datasets

This menu entry provides quick access to several MIB demo datasets and full projects for image segmentation using deep learning. The datasets are grouped by techniques used for data collection.

DeepMIB projects->Synthetic 2D large spots

A complete DeepMIB project with a synthetic dataset generated for quick tests of semantic segmentation approaches.
The dataset includes a trained DeepLabV3-Resnet18 network for detection of large spots on a black background. The network can be opened by loading "2D_LargeSpots_2cl_DeepLabV3.mibCfg" file via
Menu->Tools->Deep learning segmentation->Options tab->Config files->Load
Reference:

DeepMIB projects->Synthetic 2D small spots

A complete DeepMIB project with a synthetic dataset generated for quick tests of semantic segmentation approaches.
The dataset includes a trained U-net network for detection of small random spots of 2 colors on a black background. The network can be opened by loading "2D_SmallSpots_3cl_Unet.mibCfg" file via
Menu->Tools->Deep learning segmentation->Options tab->Config files->Load
Reference:

DeepMIB projects->Synthetic 2.5D large spots

A complete DeepMIB project with a synthetic dataset generated for quick tests of 2.5D depth-to-color semantic segmentation approaches
The dataset includes a trained 2.5D DeepLabV3-Resnet18 and 2.5D U-net networks, where the 5-slice subvolumes were used for segmentation of large 3D spots on a black background. In addition, there are 2D spots that should not be segmented.

The trained networks are

• Spots_25D_DLv3RN18_Z2C_xy200z5 - DeepLabV3-based 2.5D Depth-to-Color network with patches of 200x200x5
• Spots_25D_Unet_Z2C_xy200z5 - U-net-based 2.5D Depth-to-Color network with patches of 200x200x5

The networks can be loaded by opening their config files by

• Menu->Tools->Deep learning segmentation->Options tab->Config files->Load
• Drag and drop of the config file into DeepMIB window

Reference:

DeepMIB projects->Synthetic 2D patch-wise

A complete DeepMIB project with a synthetic dataset generated for quick tests of the patch-wise segmentation approaches
The dataset includes a trained Resnet18 network for detection of patches that belong to large white spots on black background.
The network can be opened by loading "2D_LargeSpots_Patchwise_Resnet18.mibCfg" file via
Menu->Tools->Deep learning segmentation->Options tab->Config files->Load
Reference:

DeepMIB projects->2D EM membranes

A complete DeepMIB project with a segmentation of membranes from serial-section TEM images
The dataset includes a trained U-net and DeepLabV3-Resnet18 networks for semantic segmentation. This is an updated project that has been published with the DeepMIB paper: DeepMIB: User-friendly and open-source software for training of deep learning network for biological image segmentation, Figure 1a. Check readme.txt file for details.
The network can be opened by loading "2D_EM_membranes_Unet.mibCfg" or "2D_EM_membranes_DLabRN18.mibCfg" file via
Menu->Tools->Deep learning segmentation->Options tab->Config files->Load

DeepMIB projects->2D LM nuclei

A complete DeepMIB project with a segmentation of nuclei, their boundaries and touching edges
The dataset includes a trained U-net network for semantic segmentation. This is an updated project that has been published with the DeepMIB paper: DeepMIB: User-friendly and open-source software for training of deep learning network for biological image segmentation, Figure 1b. Check readme.txt file for details.
The network can be opened by loading "valid_252px_32patches_50ep.mibCfg" file via
Menu->Tools->Deep learning segmentation->Options tab->Config files->Load

DeepMIB projects->3D EM mitochondria

A complete DeepMIB project with a segmentation of mitochondria
The dataset includes a trained 3D U-net network for semantic segmentation. This is an updated project that has been published with the DeepMIB paper: DeepMIB: User-friendly and open-source software for training of deep learning network for biological image segmentation, Figure 1c. Check readme.txt file for details.
The network can be opened by loading "NoValidation_valid_Aug_128px_256pat.mibCfg" file via
Menu->Tools->Deep learning segmentation->Options tab->Config files->Load

DeepMIB projects->3D LM inner hair cells

A complete DeepMIB project with a segmentation of inner hear cells, nuclei and synapses
The dataset includes a trained 3D anisotropic U-net network for semantic segmentation. This is an updated project that has been published with the DeepMIB paper: DeepMIB: User-friendly and open-source software for training of deep learning network for biological image segmentation, Figure 1d. Check readme.txt file for details.
The network can be opened by loading "InnerEar3D_Hybrid_Same_136x64px_120ep.mibCfg" file via
Menu->Tools->Deep learning segmentation->Options tab->Config files->Load

OMERO Import

Establish connection and load images from OMERO server. Requires OMERO server files to be downloaded. Please refer to the System Requirements pages for details of installation.

Selection of a server (please do not copy/paste the password!):

In the following dialog it is possible to select desired dataset and the range to take

A brief demonstration is available in the following video:
https://youtu.be/iR7OL0eJGuw

Batch processing...

With the batch processing mode many of the image processing operations may be designed into an image processing workflow, which may automatically applied to multiple images.

Please refer to the Batch processing... section for details.

Chopped images...

This is a special mode that allows to split a large dataset into defined number of smaller ones and combine them back later. It is also possible to fuse the previously cropped dataset to the bigger one.

Please refer to the Chopped images... section for details.

Rename and shuffle

Two tools placed under this menu entry allow to shuffle files for their blind modeling, when the user does not know which file belong to which condition. The models from the shuffled files can be converted back over the original filenames for analysis.

Please refer to the Rename and shuffle section for details.

Export image to

Export images to the main MATLAB workspace or Imaris.

- MATLAB
In addition to the image variable, another variable with parameters of the dataset (containers.Map class) is automatically created. This |containers.Map| can be imported back to |MIB| later together with the modified dataset to restore parameters of the dataset.

A brief demonstration on Import/Export is available in the following video:
https://youtu.be/zUJ1RUuTLVs

- Imaris
When Imaris is installed, the images can also be exported to Imaris.

collapse all

Save image as

Save the open dataset to a disk.

Make movie

Save dataset as a movie file. All objects that are shown in the image view window will be captured. Note! If the image width is too small the scale bar is not rendered. See more here.

Make snapshot

Make snapshot of the current slice. All objects that are shown in the image view window will be captured. See more here.

collapse all

Render volume

3D visualization of volumes is available using 3 different ways

MIB rendering

Click to open documentation about the new 3D volume viewer is available

Starting from MIB (version 2.5) and MATLAB R2018b the volumes can be directly visualized in MIB using hardware accelerated volume rendering engine. The datasets for visuzalization can be downsampled during the export. It is possible to make snapshots and animations.

MIB version 2.84 comes with the updated 3D volume rendering that allows to render volumes with 1-3 color channels together with the models. As for MIB 2.84 the new 3D viewer is only available in MIB for MATLAB.

Volume rendering engines of MIB

MIB 2.84, R2022b or newer Volumes (1-3 colors) and models	MIB 2.5, R2018b or newer only 1-channel volumes or single material models
Limitations Available only for MIB for MATLAB (at least in MIB 2.84) One volume at the time	Limitations One volume at the time; it can be image or a material of the model Only grayscaled 3D data, i.e. a single color channel Scale bar is not yet available

The version of MIB rendering engine can be selected from preferences:

MIB->Menu->File->Preferences->User interface->3D rendering engine

A brief demonstration is available in the following video:
https://youtu.be/J70V33f7bas

MATLAB Volume Viewer

When MIB is used with MATLAB R2017a or newer it is possible to export the open dataset to Volume Viewer application. Please note that this feature is not available for the compliled version of MIB

A brief demonstration is available in the following video:
https://youtu.be/J70V33f7bas

Fiji 3D viewer

Please refer to details in the Microscopy Image Browser System Requirements Fiji section for installation of Fiji.

A brief demonstration is available in the following video:
https://youtu.be/DZ1Tj3Fh2HM

Additional dialog prompts for extra parameters for rendering:

collapse all

Preferences

View and edit preferences of Microscopy Image Browser. Allows to modify colors of the Selection, Model and Mask layers, default behaviour of the mouse wheel and keys, settings of Undo. See more...

Please note, MIB stores its configuration parameters in a file that is automatically generated after closing of MIB

Back to Index --> User Guide --> Menu

Published with MATLAB® R2023b and subsequently processed by prettify_MATLAB_html V6.8b2