Academic Journal
OMEGA V2: GPU-accelerated Python and MATLAB software for PET, SPECT, and CT reconstruction.
| Τίτλος: | OMEGA V2: GPU-accelerated Python and MATLAB software for PET, SPECT, and CT reconstruction. |
|---|---|
| Συγγραφείς: | Wettenhovi VV; Department of Technical Physics, University of Eastern Finland, Kuopio, Finland.; Planmeca Group, Helsinki, Finland., Saarlemo N; Department of Technical Physics, University of Eastern Finland, Kuopio, Finland., Djurabekova N; Planmeca Group, Helsinki, Finland., Kortelainen M; Department of Technical Physics, University of Eastern Finland, Kuopio, Finland.; Diagnostic Imaging Centre, Kuopio University Hospital, Wellbeing services county of North Savo, Kuopio, Finland., Kolehmainen V; Department of Technical Physics, University of Eastern Finland, Kuopio, Finland. |
| Πηγή: | Physics in medicine and biology [Phys Med Biol] 2026 Jul 20; Vol. 71 (14). Date of Electronic Publication: 2026 Jul 20. |
| Τύπος έκδοσης: | Journal Article |
| Γλώσσα: | English |
| Στοιχεία περιοδικού: | Publisher: IOP Publishing Country of Publication: England NLM ID: 0401220 Publication Model: Electronic Cited Medium: Internet ISSN: 1361-6560 (Electronic) Linking ISSN: 00319155 NLM ISO Abbreviation: Phys Med Biol Subsets: MEDLINE |
| Imprint Name(s): | Original Publication: Bristol : IOP Publishing |
| Ιατρικοί όροι (MeSH): | Image Processing, Computer-Assisted*/methods , Software* , Computer Graphics* , Tomography, Emission-Computed, Single-Photon* , Positron-Emission Tomography* , Tomography, X-Ray Computed*, Time Factors ; Algorithms |
| Περίληψη: | Objective. OMEGA V2 is open-source software for GPU-accelerated image reconstruction in positron emission tomography (PET), single photon emission computed tomography (SPECT), and computed tomography (CT). The software offers flexible GPU accelerated image reconstruction methods and tools for imaging algorithm development which are accessible from Python, MATLAB, and GNU Octave. This paper presents the software architecture, projector models, algorithms, and demonstrates its performance with realistic high-resolution 3D examples from PET, SPECT and cone-beam CT.Approach. OMEGA V2 is based on OpenCL and CUDA allowing wide GPU support. The software provides modular forward/backprojector operators and a broad suite of built-in iterative algorithms and regularization models. It supports features such as time-of-flight imaging, list-mode reconstruction, multi-resolution approaches, and various physical corrections including attenuation, scatter, and normalization.Main results. OMEGA V2 provides a unified open-source reconstruction framework for PET, SPECT and CT, including hybrid workflows such as PET/CT and SPECT/CT. It provides cross-vendor GPU acceleration via OpenCL, supporting AMD and Intel devices alongside CUDA-capable GPUs, and introduces a new Python interface that complements and mirrors the existing MATLAB/GNU Octave workflow. The software substantially extends prior OMEGA releases with SPECT functionality, extensive CT functionality, and a Python implementation with wide interoperability such as with PyTorch. High-resolution 3D examples in PET, SPECT, and CBCT demonstrate high-quality reconstructions and fast runtimes on modern consumer GPUs.Significance. Combination of PET, SPECT, and CT in an open-source, GPU-optimized framework with broad algorithmic and projector coverage offers a unified suite for computational imaging, method development and translation of methods in CT, PET and SPECT, and their hybrid combinations (PET/CT, SPECT/CT). Its open-source nature, extensive algorithm library, and flexible programming interfaces enable users to develop custom reconstruction methods with access to GPU-accelerated projectors. (Creative Commons Attribution license.) |
| Contributed Indexing: | Keywords: CT; PET; SPECT; image reconstruction; open source; software; tomography |
| Entry Date(s): | Date Created: 20260617 Date Completed: 20260720 Latest Revision: 20260720 |
| Update Code: | 20260720 |
| DOI: | 10.1088/1361-6560/ae7ec7 |
| PMID: | 42309144 |
| Βάση Δεδομένων: | MEDLINE |
| ISSN: | 1361-6560 |
|---|---|
| DOI: | 10.1088/1361-6560/ae7ec7 |