OMEGA is an open-source MATLAB/GNU Octave Emission and Transmission Tomography Software. The purpose of OMEGA is twofold. It has been created and maintained by Ville-Veikko Wettenhovi.
OMEGA is designed primarily to allow easy, fast and efficient reconstruction of any positron emission tomography (PET) and computed tomography (CT) data, including simulated GATE data. Furthermore, it is intended for easy algorithmic development as it allows easy matrix-free implementation of the forward (
A * x) and backward (
A' * y) projections.
OMEGA is a software for MATLAB and GNU Octave to reconstruct data obtained with a positron emission tomography or CT scanner. This software also allows to easily reconstruct ASCII (text list), LMF (binary list-mode) or ROOT (enhanced list-mode) data obtained from GATE PET simulations and image-CT projections for CT. See Features section below for more information on available features and Known Issues and Limitations for software limitations. If you wish to add your own code (e.g. reconstruction algorithm) see Contributing code to OMEGA.
The algorithms implemented so far are:
- Improved Siddon’s ray tracer algorithm for the system matrix creation (code for regular Siddon available, but not used)
- Orthogonal distance-based ray tracer
- Volume of intersection ray tracer (THOR)
- Maximum Likelihood Expectation Maximization (MLEM)
- Ordered Subsets Expectation Maximization (OSEM)
- Complete-data Ordered Subsets Expectation Maximization (COSEM)
- Enhanced COSEM (ECOSEM)
- Accelerated COSEM (ACOSEM)
- Row-Action Maximum Likelihood Algorithm (RAMLA)
- Relaxed OSEM (ROSEM)
- Rescaled Block-Iterative EM (RBI)
- Dynamic RAMLA (DRAMA)
- Modified RAMLA (MRAMLA), aka modified BSREM-2
- Block Sequential Regularized Expectation Maximization (BSREM)
- One-step-late algorithm (OSL)
- Preconditioned Krasnoselskii-Mann algorithm (PKMA)
- Quadratic prior (Gibbs prior with quadratic potential function)
- Median Root Prior (MRP)
- L-filter (MRP-L) prior
- Finite Impulse Response Median Hybrid (MRP-FMH) prior
- Weighted mean prior
- Total variation (TV)
- Total generalized variation (TGV)
- Anisotropic diffusion (AD) Median Root Prior
- Asymmetric parallel levels sets prior (APLS)
- Non-local means prior (NLM), including non-local TV
- Relative difference prior
Publications introducing each of the afore-mentioned methods & features can be found in OMEGA’s Github reference page.
It is highly recommended that you visit the README and Getting Started pages first. Other practical information regarding the usage of OMEGA can be found in OMEGA’s useful information & recommendations page.
Citations: If you wish to use this software in your work, cite this paper: V-V Wettenhovi et al 2021 Phys. Med. Biol. 66 065010. The peer reviewed (open access) paper on OMEGA can be found from https://doi.org/10.1088/1361-6560/abe65f