Ensemble of neural networks for 3D position estimation in monolithic PET detectors

We propose an ensemble of multilayer feedforward neural networks to estimate the 3D position of photoelectric interactions in monolithic detectors. The ensemble is trained with data generated from optical Monte Carlo simulations only. The originality of our approach is to exploit simulations to obtain reference data, in combination with a variability reduction that the network ensembles offer, thus, removing the need of extensive per-detector calibration measurements. This procedure delivers an ensemble valid for any detector of the same design. We show the capability of the ensemble to solve the 3D positioning problem through testing four different detector designs with Monte Carlo data, measurements from physical detectors and reconstructed images from the MindView scanner. Network ensembles allow the detector to achieve a 2–2.4 mm FWHM, depending on its design, and the associated reconstructed images present improved SNR, CNR and SSIM when compared to those based on the MindView built-in positioning algorithm.

• Publication year

  2019

• Venue

  Physics in Medicine and Biology

• Publication date

  2019-10-04

• Fields of study

  Medicine, Physics, Computer Science, Engineering

• Identifiers
  DOI 10.1088/1361-6560/ab3b86PMID 31416053
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar, PubMed

• No claims are published for this paper.

• No concepts are published for this paper.

• Initial Results of the MINDView PET Insert Inside the 3T mMR

  2019cited by this paper
• Image Enhancement With PDEs and Nonconservative Advection Flow Fields

  2019cited by this paper
• CASToR: a generic data organization and processing code framework for multi-modal and multi-dimensional tomographic reconstruction

  2018cited by this paper
• A Novel DOI Positioning Algorithm for Monolithic Scintillator Crystals in PET Based on Gradient Tree Boosting

  2018cited by this paper
• Gradient Tree Boosting-Based Positioning Method for Monolithic Scintillator Crystals in Positron Emission Tomography

  2018cited by this paper
• Performance Study of a Large Monolithic LYSO PET Detector With Accurate Photon DOI Using Retroreflector Layers

  2017cited by this paper
• Sub-millimetre DOI detector based on monolithic LYSO and digital SiPM for a dedicated small-animal PET system

  2016cited by this paper
• Development of new scintillators for medical applications

  2016cited by this paper
• Experimental Validation of an Efficient Fan-Beam Calibration Procedure for $k$-Nearest Neighbor Position Estimation in Monolithic Scintillator Detectors

  2015cited by this paper
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

  2015cited by this paper
• Detector block based on arrays of 144 SiPMs and monolithic scintillators: A performance study

  2015cited by this paper
• High-volume silicon photomultiplier production, performance, and reliability

  2014cited by this paper
• Dropout: a simple way to prevent neural networks from overfitting

  2014cited by this paper
• High performance detector head for PET and PET/MR with continuous crystals and SiPMs

  2013cited by this paper
• 3D position estimation using an artificial neural network for a continuous scintillator PET detector

  2013cited by this paper
• Improved Nearest Neighbor Methods for Gamma Photon Interaction Position Determination in Monolithic Scintillator PET Detectors

  2011cited by this paper
• Adaptive Subgradient Methods for Online Learning and Stochastic Optimization

  2011cited by this paper
• Rectified Linear Units Improve Restricted Boltzmann Machines

  2010cited by this paper
• Optical simulation of monolithic scintillator detectors using GATE/GEANT4

  2010cited by this paper
• Characterization of a PET detector head based on continuous LYSO crystals and monolithic, 64-pixel silicon photomultiplier matrices

  2010cited by this paper
• Monolithic scintillator PET detectors with intrinsic depth-of-interaction correction

  2009cited by this paper
• A novel, SiPM-array-based, monolithic scintillator detector for PET

  2009cited by this paper
• Revisited position arithmetics for LaBr3:Ce continuous crystals

  2009cited by this paper
• Evaluation of Machine Learning Algorithms for Localization of Photons in Undivided Scintillator Blocks for PET Detectors

  2008cited by this paper
• Neural network-based position estimators for PET detectors using monolithic LSO blocks

  2004cited by this paper
• Image quality assessment: from error visibility to structural similarity

  2004cited by this paper
• Geant4—a simulation toolkit

  2003cited by this paper
• A more physical approach to model the surface treatment of scintillation counters and its implementation into DETECT

  1996cited by this paper
• Approximation capabilities of multilayer feedforward networks

  1991cited by this paper
• Neural Network Ensembles

  1990cited by this paper
• Multilayer feedforward networks are universal approximators

  1989cited by this paper
• Surface Reflection: Physical and Geometrical Perspectives

  1989cited by this paper
• [Scintillation camera].

  1967cited by this paper

• Stacking ensemble machine learning for predicting photodetector performance under varying illumination intensities

  2026influential citation
• Design and Validation of a SPECT Prototype for Treatment Monitoring in BNCT and First Experimental Tomographic Results

  2026cites this paper
• Machine learning positioning algorithms for long semi-monolithic scintillator PET detectors

  2025cites this paper
• Depth-of-interaction encoding techniques for pixelated PET detectors enabled by machine learning methods and fast waveform digitization

  2025cites this paper
• Gamma-Ray Position-of-Interaction Estimation in a Thick Monolithic LaBr3 Detector Using Artificial Neural Networks

  2025cites this paper
• Improving timing resolution of BGO for TOF-PET: a comparative analysis with and without deep learning

  2025cites this paper
• Accurate and automated energy calibration for pixelated scintillator detectors through a convolution neural network

  2025cites this paper
• CRYSP: a total-body PET based on cryogenic cesium iodide crystals

  2025cites this paper
• Performance study of monolithic crystal detector based on neighboring crystal light sharing with dual-ended readout

  2025cites this paper
• A High-Resolution Portable Gamma-Camera for Estimation of Absorbed Dose in Molecular Radiotherapy

  2024cites this paper
• A deep neural network for positioning and inter-crystal scatter identification in multiplexed PET detectors: a simulation study

  2024cites this paper
• The quest for multifunctional and dedicated PET instrumentation with irregular geometries

  2023cites this paper
• Single-Ended Readout Depth-of-Interaction Measurements Based on Random Forest Algorithm

  2023cites this paper
• A CNN-based four-layer DOI encoding detector using LYSO and BGO scintillators for small animal PET imaging

  2023cites this paper
• Angular Irradiation Methods for DOI Calibration of Light-Sharing Detectors—A Perspective for PET In-System Calibration

  2023cites this paper
• A simulation study of 1D U-Net-based inter-crystal scatter event recovery of PET detectors

  2023cites this paper
• A finely segmented semi‐monolithic detector tailored for high‐resolution PET

  2023cites this paper
• A study on a nonlinear least-squares fitting method for 3D positioning of gamma rays based on monolithic crystal and SiPM array

  2022cites this paper
• A Semi-Monolithic Detector providing intrinsic DOI-encoding and sub-200 ps CRT TOF-Capabilities for Clinical PET Applications.

  2022cites this paper
• A neural network-based algorithm for simultaneous event positioning and timestamping in monolithic scintillators

  2022cites this paper
• Simulation study on 3D convolutional neural networks for time-of-flight prediction in monolithic PET detectors using digitized waveforms

  2022cites this paper
• A mobile high-resolution gamma camera for therapeutic dose control during radionuclide therapy

  2022cites this paper
• Calculation of the Spatial Resolution of a Positron-Emission Tomograph Element Based on a Crystalline Plate Using Neural Networks

  2022cites this paper
• Reducing Calibration Time in PET Systems Based on Monolithic Crystals

  2021cites this paper
• Depth of Interaction Measurements Based on Rectangular Light Sharing Window Technology and Nine-Crystals-to-One-SiPM Coupling Method

  2021cites this paper
• High-performance coded aperture gamma camera based on monolithic GAGG:Ce crystal.

  2021cites this paper
• Artificial neural networks for positioning of gamma interactions in monolithic PET detectors

  2021cites this paper
• Advanced Monte Carlo simulations of emission tomography imaging systems with GATE

  2021cites this paper
• Event Localization in Continuous Crystal Scintillation Cameras using Distribution Matching Neural Networks

  2021cites this paper
• Evolution of PET Detectors and Event Positioning Algorithms Using Monolithic Scintillation Crystals

  2021cites this paper
• Machine Learning aided 3D-position reconstruction in large LaCl3 crystals

  2021cites this paper
• Deep residual-convolutional neural networks for event positioning in a monolithic annular PET scanner

  2021cites this paper
• High throughput software-based gradient tree boosting positioning for PET systems

  2021cites this paper
• High-resolution monolithic LYSO detector with 6-layer depth-of-interaction for clinical PET

  2021cites this paper
• Inter-crystal scattering recovery of light-sharing PET detectors using convolutional neural networks

  2021cites this paper
• Artificial intelligence with deep learning in nuclear medicine and radiology

  2021cites this paper
• Characterization of a PET Detector based on Semi-Monolithic Crystals

  2021cites this paper
• A preliminary study on 3D position reconstruction of monolithic crystal readout

  2020cites this paper
• Convolutional Neural Network for Crystal Identification and Gamma Ray Localization in PET

  2020cites this paper
• A preliminary study on 3D position reconstruction of monolithic crystal readout

  2020cites this paper
• Sub-2 mm Depth of Interaction Localization in PET Detectors with Prismatoid Light Guide Arrays and Single-Ended Readout using Convolutional Neural Networks.

  2020cites this paper
• Machine Learning aided 3D-position reconstruction in large LaCl$_{3}$ crystals

  2020cites this paper
• Characterization and Simulation of an Adaptable Fan-Beam Collimator for Fast Calibration of Radiation Detectors for PET

  2020cites this paper