Related Experiment Video
Updated: Jul 3, 2026

Cortical Bone Assessment Using Ultrasonic Guided Waves: A Reproducibility Study in a Healthy Population
Published on: January 31, 2025
Dynamic PET reconstruction using wavelet regularization with adapted basis functions
Jeroen Verhaeghe1, Dimitri Van de Ville, Ildar Khalidov
1Department of Electronics and Information Systems, MEDISIP, Ghent University-IBBT-IBiTech, De Pintelaan 185 block B, B-9000 Ghent, Belgium. jeroen.verhaeghe@ugent.be
This study introduces exponential-spline wavelets for dynamic positron emission tomography (PET) reconstruction, improving image quality and reducing patient radiation exposure. These novel wavelets enhance signal-to-noise ratio and sparsity in PET imaging.
Area of Science:
- Medical Imaging
- Signal Processing
- Computational Science
Background:
- Positron emission tomography (PET) reconstruction is an ill-posed problem requiring regularization techniques.
- L(1)-regularization in the wavelet domain promotes sparse solutions, efficiently implemented via iterative algorithms.
- Dynamic PET data analysis necessitates spatio-temporal reconstruction methods.
Purpose of the Study:
- To extend iterative L(1)-regularization for dynamic PET data reconstruction.
- To introduce and evaluate exponential-spline (E-spline) wavelets for modeling temporal activity curves (TACs) in PET.
- To compare the performance of E-spline wavelets against conventional wavelets in PET image reconstruction.
Main Methods:
- Application of an iterative L(1)-regularization algorithm to dynamic PET data.
- Development and integration of E-spline wavelets, tailored for PET TACs, into the reconstruction process.
- Evaluation through 1-D TAC fitting, simulated, and clinical PET data reconstruction, analyzing signal-to-noise ratio (SNR) and wavelet coefficient sparsity.
Main Results:
- E-spline wavelets naturally model tracer distribution dynamics and arise from compartmental models.
- Optimal E-spline parameters were investigated for their impact on reconstruction quality.
- E-spline wavelets demonstrated superior performance over conventional Battle-LemariE wavelets in terms of SNR and sparsity.
Conclusions:
- Spatio-temporal regularization with E-spline wavelets significantly improves dynamic PET reconstruction.
- E-spline wavelets enable equivalent reconstruction quality with a 40% reduction in detected coincidences.
- This translates to enhanced image quality for the same data acquisition or reduced patient radiation dose for equivalent image quality.
Related Concept Videos
Partial Differential Equations
Reconstruction of Signal using Interpolation
Wave Parameters
