A Variational Bayesian inference method for parametric imaging of PET data

Castellaro, Marco; Rizzo, Gaia; Tonietto, Mattéo; Veronese, Mattia; Turkheimer, Federico; Chappell, Michael A.; Bertoldo, Alessandra

doi:10.1016/j.neuroimage.2017.02.009

Cited by 26 publications

(38 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This formula associates the calculation of the posterior distribution of the factors specifying the data and the model (p(θ\y, m)) to the a priori distributions of the factors to be predicted (p(θ\m)). These distributions are acquired through the likelihood (p(y\θ, m)) and the probability density function, which defines the data specifying the factors and the model (Castellaro et al 2017). Model indication can be measured as a normalization constant for the product of the likelihood of the data and the prior probability of the factors.…”

Section: Modellingmentioning

confidence: 99%

See 1 more Smart Citation

Brain activity detection in single- and multi-subject PET data by Bayesian analysis

Aljobouri

2020

Res. Biomed. Eng.

View full text Add to dashboard Cite

Purpose Positron emission tomography (PET) is a functional neuroimaging method that maps brain activity non-invasively. Statistical methods play an essential part in understanding and analysing functional PET data. Several Bayesian approaches have been proposed for neuroimaging techniques that arrange information on the brain structure or activity function. In this paper, Bayesian analysis was used to detect functional brain activity in single-and multiple-subject PET data. Methods Free PET dataset analyses for a single subject and five multiple subjects were conducted. A total of 72 functional PET images were processed, 12 for each one of the five multiple subjects and for the single subject. Several ways to design multiplesubject PET data were introduced in this work. Bayesian analysis was performed on the five multiple-subject and the singlesubject PET data. A comparison was presented to determine which statistical matrix design is applicable for brain detection activity in PET data. Results The results of the design matrix and brain activity detection were presented for each selected design matrix. The Bayesian estimation of each case of the PET dataset for all the subjects was plotted. The brain activity was plotted as voxels on a transparent brain image in three orthogonal planes. The voxels were visualised using the maximum intensity projection method. Conclusion Results showed that brain activity could not be detected easily in single-subject PET data. Finding the activity in multiple subjects depended on the design matrix used for PET data analysis.

show abstract

Section: Modellingmentioning

confidence: 99%

“…Another study applied the variational Bayesian (VB) method to PET dataset for the first time (Castellaro et al 2017). VB was adjusted to the non-uniform noise distribution of the PET dataset.…”

Section: Introductionmentioning

confidence: 99%

Brain activity detection in single- and multi-subject PET data by Bayesian analysis

Aljobouri

2020

Res. Biomed. Eng.

View full text Add to dashboard Cite

show abstract

“…Given the observed data Q and the parameter set Θ , which contains n independent model parameters, Θ 1 ,Θ 2 ...Θ n , the posterior of Θ (Castellaro et al, ) can be expressed as

p false(normalΘ false| Q false) = \frac{p false(Q false| normalΘ false) p false(normalΘ false)}{p false(Q false)}

…”

Section: Proposed Modelmentioning

confidence: 99%

“…Given the observed data Q and the parameter set , which contains n independent model parameters, Θ 1 , Θ 2 ...Θ n , the posterior of (Castellaro et al, 2017) can be expressed as…”

Section: Optimizationmentioning

confidence: 99%

Linear Multistep F10.7 Forecasting Based on Task Correlation and Heteroscedasticity

Wang

Zhong

et al. 2018

Earth and Space Science

View full text Add to dashboard Cite

Solar 10.7‐cm radio flux (F10.7) is an important measure of solar radio emission activity. Accurate F10.7 forecasting plays a key role in both space weather and global environment forecasts. We discover that forecasting errors are heteroscedastic, something that is not often considered in previous models. In addition, task correlation between different forecasting steps is ignored in current multistep‐ahead forecast models. In this work, we propose a linear multistep forecasting model based on the correlation between different forecasting steps and the characteristic of heteroscedasticity. Further, we introduce a variational Bayesian procedure to optimize the model. The performance of the proposed model is tested on F10.7 historical data. The results show that the proposed model improves the performance of multistep F10.7 forecasting by considering correlation and heteroscedasticity.

show abstract

“…Such models, like Bayesian classifier, linear discriminant analysis, binary logistic regression, adaptive neurofuzzy inference system, etc., were successfully applied to skin detection [7,[22][23][24][25][26]. Among them, the Bayesian classifier is especially noteworthy not only in the field of skin detection but also in other disciplines because it provides the information concerning the probability that an observation belongs to a class, thereby evaluating the reliability of the result [27][28][29]. However, the Bayesian classifier, as well as other methods, still suffers from low performance, especially the high false detection rate (the percentage of nonskin classified as skin).…”

Section: Introductionmentioning

confidence: 99%

A New Efficient Approach to Detect Skin in Color Image Using Bayesian Classifier and Connected Component Algorithm

Nguyen‐Trang

2018

Mathematical Problems in Engineering

View full text Add to dashboard Cite

Skin detection is an interesting problem in image processing and is an important preprocessing step for further techniques like face detection, objectionable image detection, etc. However, its performance has not really been high because of the high overlapped degree between "skin" and "nonskin" pixels. This paper proposes a new approach to improve the skin detection performance using the Bayesian classifier and connected component algorithm. Specifically, the Bayesian classifier is utilized to identify "true skin" pixels using the first posterior probability threshold, which is approximate to 1, and to identify "skin candidate" pixels using the second posterior probability threshold. Subsequently, the connected component algorithm is used to find all the connected components containing the "skin candidate" pixels. According to the fact that a skin pixel often connects with other skin pixels in an image, all pixels in a connected component are classified as "skin" if there is at least one "true skin" pixel in that connected component. It means that the "nonskin" pixels whose color is similar to skin are classified as "nonskin" when they have the posterior probabilities lower than the first posterior probability threshold and do not connect with any "true skin" pixel. This idea can help us to improve the skin classification performance, especially the false positive rate.

show abstract

A Variational Bayesian inference method for parametric imaging of PET data

Cited by 26 publications

References 41 publications

Brain activity detection in single- and multi-subject PET data by Bayesian analysis

Brain activity detection in single- and multi-subject PET data by Bayesian analysis

Linear Multistep F10.7 Forecasting Based on Task Correlation and Heteroscedasticity

A New Efficient Approach to Detect Skin in Color Image Using Bayesian Classifier and Connected Component Algorithm

Contact Info

Product

Resources

About