Projects – Work
Mobius3D
The Complete Patient QA System
The Mobius3D® software platform includes integration modules for quality assurance (QA) of every patient plan, every plan delivery, and every patient cone-beam CT (CBCT).
It features an efficient workflow which relies on measurements and images already generated during treatment workflows, with automatic analysis and no external hardware requirements.
A single server can operate all platform modules for up to 10 linear accelerators and has storage available for up to 20,000 patients.
Mobius3D receives all your patients’ treatment plans via DICOM-RT and generates results within an intuitive web interface.
With its integrated patient and machine QA modules, the platform provides six layers of error detection:
Machine QA (DoseLab)
3D patient plan QA
3D IMRT/VMAT pre-treatment QA
Online patient positioning QA
Quantitative anatomy verification
3D daily treatment QA
Self-developed implants and surgical instruments for shoulder (reverse shoulder arthroplasty), leg (bunion deformity correction) and knee (total knee arthroplasty) surgeries.
A grant was awarded under the Cooperation for Competitiveness and Excellence programme, within the framework of Operational Programme for Economic Development and Innovation R&F, (GINOP – 2.2.1-15).
The project was realised with effective cooperation between the University of Szeged, Sanatmetal Kft. (consortium leader) and dicomLAB Kft. All the members of the consortium are internationally recognised.
Timeframe: 09/2016 - 02/2019
Country: Hungary
Organisation: Sanatmetal Kft.
Fund: European Regional Development Fund
Projects – University
Feature selection methods and machine learning (like decision trees, neural networks, floating search methods) are robust and effective tools to detect important attributes of a given object, or to classify objects of the same type with the aid of their attributes. In discrete image reconstruction problems one can think of the projection components as the attributes of the discrete set.
In this scenario – in order to facilitate the reconstruction – the task is to determine structural or geometrical properties of the discrete set from the attributes (e.g., from the projection components). This way, a pre-processing step can decide which reconstruction algorithm to choose and how to parameterize it.
Additionally, we can investigate which are the most important projection directions or projection components that can guarantee geometrical properties. Observations of this kind can also yield characterization results, and reconstruction algorithms can be designed to directly use the information gained from the attributes by applying techniques mentioned above.
Statistical shape analysis using the Point Distribution Model
Presented the MSc student project within the scope of the Image Processing Special Programme course titled „Image Registration”, lectured by Dr. Attila Tanács.
The aim was to get familiar with the basic concepts and techniques of statistical analysis of shapes and to use Principal Component Analysis for modeling shape variation. Since the literature on these topics is quite rich, the present documentation does not contain minute details, instead presents the listed methods and the chosen approach through a thorough example. The development environment is MATLAB, which has plenty data analysis options, image processing tools and easy data visualization, among other things.
This documentation is organized in a step‐by‐step fashion, presenting the theory along with the required MATLAB code to solve the actual task described for a given step. In case several possible solutions and approaches exist to a sub-problem, a reference is made to the literature detailing other options, latter entries can be found in the bibliography.