Direct Estimation of Voxel-Wise Neurotransmitter Response Maps From Dynamic PET Data

Angelis, Georgios I.; Gillam, J. E.; Ryder, William J.; Fulton, Roger; Meikle, Steven R.

doi:10.1109/tmi.2018.2883756

Cited by 20 publications

(24 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For these simulations, the following parameters were used: t d = 20 min, t p = 25 min, α = 2, baseline release Basal = 100 nM and peak dopamine level ranged between 0% and 300% of the baseline level. See [12] for further details of kinetic parameters and input functions used in the simulation of TACs. The data were labeled with the ground-truth (activation or nonactivation) and used to train the algorithms.…”

Section: A Simulated Training Datamentioning

confidence: 99%

“…2) 2-D+Time Data: Computer simulations were again used to generate dynamic data which resembled a [ 11 C]raclopride study of the rat brain using the complete ntPET model [2] (see [12] for detailed description of kinetic parameters and input functions used in the simulation of TACs). A reference region, representing a region without specific binding was also simulated, in this case in the cerebellum.…”

Section: A Simulated Training Datamentioning

confidence: 99%

“…3) 4-D Data: 4-D GATE data based on a [ 11 C]raclopride study were generated again using the complete ntPET model [2] and kinetic parameters as specified in [12]. A coronal slice of the 3-D phantom which cuts through the striatum is shown in Fig.…”

Section: A Simulated Training Datamentioning

confidence: 99%

“…To investigate whether the lower noise in the TACs provided by direct reconstruction benefitted the MLAs, the algorithms were incorporated into the reconstruction process. In this article, we used a direct reconstruction algorithm, based on the expectation maximization framework, which decouples at each iteration the image reconstruction and parameter estimation steps [12], [17]. At every iteration, we first reconstruct the voxel-wise 4-D dynamic data and then we fit each and every voxel with the preferred kinetic model (lp-ntPET or MRTM) based on a decision metric, i.e., the classification outcome of the neural network.…”

Section: Reconstruction Algorithmsmentioning

confidence: 99%

See 3 more Smart Citations

Classification of Neurotransmitter Response in Dynamic PET Data Using Machine Learning Approaches

Fuller

Angelis

Meikle

2020

IEEE Trans. Radiat. Plasma Med. Sci.

Self Cite

View full text Add to dashboard Cite

Dynamic positron emission tomography (PET) imaging can be used to quantify changes in synaptic concentrations of endogenous neurotransmitters during cognitive tasks or pharmacological interventions. Existing pharmacokinetic models, such as the linear parametric neurotransmitter PET (lp-ntPET) method, can be used to model the measured dynamic data and characterize small transient changes in neurotransmitter levels. Application of these models to the voxel level is challenging, however, due to the high levels of noise in dynamic data, leading to a high number of false-positive responses (i.e., low specificity). In this article, we investigated the suitability of machine learning algorithms (MLAs), including support vector machine (SVM) classifiers, shallow feedforward neural networks, convolutional neural networks (CNNs) and long short-term memory (LSTM) networks, to detect and classify transient changes in voxel-wise time-activity curves. We also investigated whether the reconstruction framework (post versus direct reconstruction) had any impact on the performance of the MLAs. We used computer simulations to generate dynamic PET data, representing a [ 11 C]raclopride study, with known activation responses, across a wide range of noise levels. Different simulated data sets were used to train and test the MLAs across a range of noise levels and activation response magnitudes. Results showed the MLAs offered a large improvement in specificity without a corresponding decrease in sensitivity across all noise levels tested compared to direct application of the lp-ntPET model. They also offered a modest benefit over the currently accepted method (statistical F-test combined with cluster size analysis), for both 2-D+time data when incorporated within direct or post-reconstruction frameworks, and 4-D GATE data. Index Terms-Convolutional neural networks (CNNs), direct reconstruction, feedforward neural network, ligand displacement long short-term memory (LSTM), machine learning, neurotransmitter response, parametric imaging. I. INTRODUCTION D YNAMIC positron emission tomography (PET) imaging can be used to quantify transient changes in synaptic concentrations of endogenous Manuscript

show abstract

Section: A Simulated Training Datamentioning

confidence: 99%

Section: A Simulated Training Datamentioning

confidence: 99%

Section: A Simulated Training Datamentioning

confidence: 99%

Section: Reconstruction Algorithmsmentioning

confidence: 99%

See 2 more Smart Citations

Classification of Neurotransmitter Response in Dynamic PET Data Using Machine Learning Approaches

Fuller

Angelis

Meikle

2020

IEEE Trans. Radiat. Plasma Med. Sci.

Self Cite

View full text Add to dashboard Cite

show abstract

“…The model has been successfully applied to 11 C-raclopride data to estimate the temporal dynamics of dopamine release in the mesolimbic circuit during smoking (Cosgrove et al, 2014;Kim et al, 2014). Recently, lp-ntPET has been applied in preclinical 11 C-raclopride PET studies following amphetamine administration (Angelis et al, 2019).…”

Section: Temporal Changes In Neurotransmitter Levels: Models With Timmentioning

confidence: 99%

Methods for Quantifying Neurotransmitter Dynamics in the Living Brain With PET Imaging

et al. 2020

View full text Add to dashboard Cite

Positron emission tomography (PET) neuroimaging in neuropsychiatry is a powerful tool for the quantification of molecular brain targets to characterize disease, assess disease subtype differences, evaluate short-and long-term effects of treatments, or even to measure neurotransmitter levels in healthy and psychiatric conditions. In this work, we present different methodological approaches (time-invariant models and models with time-varying terms) that have been used to measure dynamic changes in neurotransmitter levels induced by pharmacological or behavioral challenges in humans. The developments and potential use of hybrid PET/magnetic resonance imaging (MRI) for neurotransmission brain research will also be highlighted.

show abstract

Clinically Translatable Direct Patlak Reconstruction from Dynamic PET with Motion Correction Using Convolutional Neural Network

Xie

Gong

Guo

et al. 2020

Medical Image Computing and Computer Assisted Intervention – MICCAI 2020

View full text Add to dashboard Cite

Patlak model is widely used in 18 F-FDG dynamic positron emission tomography (PET) imaging, where the estimated parametric images reveal important biochemical and physiology information. Because of better noise modeling and more information extracted from raw sinogram, direct Patlak reconstruction gains its popularity over the indirect approach which utilizes reconstructed dynamic PET images alone. As the prerequisite of direct Patlak methods, raw data from dynamic PET are rarely stored in clinics and difficult to obtain. In addition, the direct reconstruction is time-consuming due to the bottleneck of multiple-frame reconstruction. All of these impede the clinical adoption of direct Patlak reconstruction. In this work, we proposed a data-driven framework which maps the dynamic PET images to the high-quality motioncorrected direct Patlak images through a convolutional neural network. For the patient's motion during the long period of dynamic PET scan, we combined the correction with the backward/forward projection in direct reconstruction to better fit the statistical model. Results based on fifteen clinical 18 F-FDG dynamic brain PET datasets demonstrates the superiority of the proposed framework over Gaussian, nonlocal mean and BM4D denoising, regarding the image bias and contrast-to-noise ratio.

show abstract

Direct Estimation of Voxel-Wise Neurotransmitter Response Maps From Dynamic PET Data

Cited by 20 publications

References 33 publications

Classification of Neurotransmitter Response in Dynamic PET Data Using Machine Learning Approaches

Classification of Neurotransmitter Response in Dynamic PET Data Using Machine Learning Approaches

Methods for Quantifying Neurotransmitter Dynamics in the Living Brain With PET Imaging

Clinically Translatable Direct Patlak Reconstruction from Dynamic PET with Motion Correction Using Convolutional Neural Network

Contact Info

Product

Resources

About