Funded by the FFG under Project 846056 (2014-2017).

Z. Bardosi, Y. Özbek, C. Plattner, W. Freysinger

The technical concept and core area of AROSCOPE involves: two dimensional (2D) visualization and spatial viewing of the patient images, calibration of the stereo cameras that mounted on the head of microscope, hand-eye calibration, patient-image registration, three dimensional (3D) segmentation, intraoperative 3D surgical navigation and overlaying of segmented anatomical structures into the ocular of the microscope in real-time.

This project provides the possibility to enhance the surgeon’s ability for a better intraoperative orientation by overlaying 3D information directly in the FOV.

Goal
The aim of this project was developing a surgical navigation and stereoscopic augmented reality concept for the use in minimal invasive microscopic image-guided surgery in ENT operations.

The concept serves user-friendly interactions and all essential steps from planning until overlaying phase and fulfills important requirements for a successfully surgery in minimal invasive interventions with navigated surgical microscope. To enhance the ability of the surgeon during the surgery and facilitate the whole complex workflow, AROSCOPE offers step by step processing of major requirements and clear guideline for the user. Consequently this system enhances the capabilities of the surgeon visual system through the combination of computer generated graphics, computer vision and advanced user interaction technology.

Screenshots

Funded by the Jubilee Fund of the Austrian National Bank, project 13003 (2008 – 2012, ongoing).

Detailed information can be found in Downloads and Publications pages.

Y. Özbek, W. Freysinger

Goal
The goal of this project is as first, represent a template (demo application) using the CTK plug-in framework that facilitates and simplifies the creation of complex applications in the biomedical field in many ways, e.g. implementation and management and as second, set up a tutorial that provides support to the development with the CTK plug-in framework.

To validate the use of CTK in computer-assisted surgery, and to facilitate the entry into the framework, a demo application is developed. It enables the loading of DICOM images, the integration of tracking data, the marker-based registration of CT data, navigation and the display of (Stereo-) video streams.

 Screenshots

Video

[There is a video that cannot be displayed in this feed. Visit the blog entry to see the video.]

Detailed information can be found in Downloads and Publications pages.

F. Gangelberger, Y. Özbek, W. Freysinger

Goal
The goal of the project is to develop an algorithm to identify/detect in the patient inserted and in the near of the surgical area positioned titanium fiducials in computed tomography (CT) images automatically, to calculate their centroids, locations, and relationships to each other, in order to realize the registration of CT images with the patient for the surgical navigation in preoperative phase without any user error.

Automatic Fiducial Detection Tool enables detection of searched objects in the background in seven steps. First is defined using the user an unique working area “region of interest (ROI)” in 3D multiplane view images that contains desired objects in it and thus by ROI reduce the processing time of the algorithm. In the second step, a 3D binary image is generated by using of the binary threshold method, which is used later for the detection, localization and measurements of geometric features of labeled objects. After segmentation or rather binarization, a structured 3D binary ball element is created based on the radius of searched objects that used in the next step in the morphological process. In the fourth step, a morphological opening filter on the binary image is applied with the previously created element in order to determine whether the object in binary image corresponds to the predefined object element. After the elimination of unrequested objects (faults) in the image, a geometry filter on the output image is applied, that is generated after morphological step, to determine and calculate the geometrical properties (centroid-coordinates in cartesian coordinate system, major/minor axes lengths, number of detected objects, elongation, eccentricity, etc.) of detected objects. In the sixth step, a distance map based on found geometric properties is drawn, that represents and simulates the relations of the localized objects, e.g. distances between objects. At the end of the process the detected objects are segmented and visualized as an iso-surface or rather 3D image.

Screenshots

Detailed information can be found in the Downloads and Publications pages.

Y. Özbek, W. Freysinger

The user has the possibility to enter the number of iterations, the number of control points and the maximum arithmetic mean value between two iterations manually and to register point clouds with the entered values iteratively. After each registration the root-mean-square error (RMS) and a transformation matrix is computed and stored separately with registration result (registered model) in poly file format.

Screenshots

Y. Özbek, W. Freysinger

Goal 
The goal of this project was, implementing an open-source computer program that in the first step reads DICOM formatted 2D CT images, then represents images in 2D axial, sagittal, coronal, and 3D multiplane view. In the second step, the loaded images are visualized and processed in different views at different windows, such as zoom in/out, changing of the position of a view in the window, windowing with Hounsfield-scale, scrolling of images, changing of opacity, changing of color, point selection, screenshot, rotating of image, list specific information from the patient and CT images, such as patient name, number of pictures, picture size, etc.

In the third step, the 2D images should be visualized in a larger window as Cine Mode and scrolled forward and backward for all views automatically. The Cine Mode should be used for enlarged representations of 2D images and 3D models.

In the next step, 2D images should be converted by specific algorithms (volume rendering, watershed segmentation and region-growing segmentation) in a 3D Model/Object. Before segmentation the program should be able the fact that the user selects itself on 2D images a region (cutting area) and afterwards segmentation algorithm should be selected, thus the processing time is reduced and the algorithm will save time.

Finally segmented 3D Model should be stored as a file in vtk-data format, so that it canb e processed or regarded in another medical computer programs.

Screenshots

Features

1. Easy and user-friendly handling
2. Visualization of the loaded 2D-images as axial, sagittal, coronal, multiplane, and 3D view (DICOM representation)
3. Image viewing for 2D images by using of mouse buttons: zoom in/out, image positioning of all views at separate windows, windowing with Hounsfield scale, rotating of 3D model,  point selection of 2D images to display all the views of the same 2D slice image, region selection of 2D images for the cutting function, scrolling
4. Image processing for 2D Images: Changing of opacity.
5. Image processing for 3D Model: Rotating and changing the color of model
6. List of specific image and patient information
7. Input field for specific values for Hounsfield scale, marching cubes volume rendering, watershed segmentation and region growing segmentation
8. Orientation with labels and a 3D orientation-model by the processing and viewing of 2D images and 3D models
9. Cine Mode for forward and backward scrolling of 2D images automatically
10. Larger representation of all views in a separate window
11. Screenshot for all views
12. Exporting of 3D models as vtk-data file
13. Live log function for user actions and program events
14. Help window
15. Loading the CT images with different sizes (512×512, 256×256, etc.)
16. Preview function before loading the CT images

Y. Özbek, W. Freysinger

Goal
The goal of this project was, developing an open-source computer program that reads the CT images, then segments those with marching cubes algorithm as 3D models and saves the segmented 3D model as vtk-data. It should be realized about read 2D images and segmented 3D models different processing and operations. As basis materials are used real CT images, which initially identifies the bone or soft tissue structures and displays 2D images with the Hounsfield scale.

In another step, read 2D images should be visualized as 3D model by using the  marching cubes algorithm, which segments the images based on predefined Hounsfield scale parameters. In last step, should the segmented and represented 3D Models should be saved. Before segmentation or saving of the images, different image processing operations and image viewing tasks can be performed.

Screenshots

Features

1. Easy and user-friendly handling
2. Visualization of the loaded 2D-images as axial, sagittal, coronal, multiplane, and 3D model (DICOM presentation)
3. Image viewing for 2D images by using of mouse buttons: zoom in/out, image positioning of all views in separate windows, windowing with Hounsfield scale, rotating of 3D model,  point selection of 2D images to display all the views of the same 2D slice image, region selection of 2D images for the cutting function, scrolling
4. Image processing for 2D Images: Changing of opacity.
5. Image processing for 3D Model: Rotating and change color of Model
6. List of specific image and patient information
7. input field to insert of specific values​​for Hounsfield scale and marching cubes
8. Cine Mode for forward and backward scrolling of 2D images automatically
9. Larger representation of all views at a separate window
10. Screenshot for all views
11. Exporting of 3D models as vtk-data file
12. Live log function for user actions and program events
13. Help window
14. Display of 2D images and 3D model on a window at the same time

Detailed information can be found in the Download page.

Y. Özbek, W. Freysinger

GantryTilt0 is implemented on Linux (Fedora 15) with programming language C++ and can be used and further expanded by integrating certain medical open-source/cross-platform libraries on Windows, Mac OS X,  etc.

Y. Özbek, W. Freysinger

The keywords are supported by the compiler mimimiC predefined as follows:

The output language which is accepted by mimimiC Virtual Machine:

Features

1. Primitive data types und complex data types (int, char, struct, array(int, char))
2. Control structures (if, else, while) with arbitrary interleaving
3. Type definitions/Structs (typedef typname aliasname; , typedef struct { int re, int im; } complx; )
4. Arithmetical and logical operators (+, -, *, /, %, ==, !=, >, < , >=, <= ||, &&)
5. Combining of brackets for arbitrary infix notation
6. Error handling for the detection of syntax errors in the source-code
7. Arrays (one Dimensional)
8. Pointer(&)
9. Functions (with Call-by-Value and Call-by-Reference)
10. File I/O (Is realized by the VM, plus there is the FOP (open), FCL (close), FRD (read), FWR (write))
11. Separate compilation
12. Symbol table
13. EBNF
14. Heap – Linker
15. Object-File generation
16. Virtual Machine
17. RISC code generation

Detailed information can be found in the Download page.

Y. Özbek, M. Kleber