luc eglin - Academia.edu (original) (raw)
Papers by luc eglin
HAL (Le Centre pour la Communication Scientifique Directe), 1999
Ce travail propose une nouvelle approche de restauration 2D des images scintigraphiques. L'or... more Ce travail propose une nouvelle approche de restauration 2D des images scintigraphiques. L'originalite repose sur la prise en compte de l'energie des photons detectes, grâce a un modele physique realiste et rapide du processus de formation des images. Le modele fournit non pas une, mais une serie d'images indexee en energie, simulant et exploitant les cameras fonctionnant en mode liste energetique. Le modele incorpore la diffusion Compton d'ordre 1, l'attenuation inhomogene par les tissus, et les effets engendres par la camera. Il permet une correction conjointe de ces effets, sans chercher a " eliminer " mais plutot a exploiter les photons diffuses. Le modele est valide par des simulations de Monte Carlo d'ordre 1 et par des mesures effectuees en environnement hospitalier. Son domaine energetique de validite est precise. Cette approche multidimensionnelle ouvre la voie a de nouvelles approches de restauration. Dans un premier temps, une analyse en ...
A method for processing a noisy digital time signal yk of digital pitch k, corresponding to an in... more A method for processing a noisy digital time signal yk of digital pitch k, corresponding to an initial analogue signal st after being conditioned by a conditioning chain. The initial analogue signal st includes at least one pulse representing information concerning at least one radiation from a radiation source, the radiation and the pulse having an energy distribution. The method includes the determination of a non-noisy digital estimation signal sk from the noisy time signal yk by using a state model representing the conditioning imposed by the conditioning chain and in that the state model includes a Markovian variable rk to be estimated whereof at least two values are associated with physical characteristics of at least two typical pulses constituting a possible representation, at least approximate, of the pulse in the signal st.
Medical Imaging 1999: Physics of Medical Imaging, 1999
ABSTRACT This paper proposes a realistic joint energy-spatial description of the various phenomen... more ABSTRACT This paper proposes a realistic joint energy-spatial description of the various phenomena encountered in low energy gamma imagery: Compton diffusion, photoelectric absorption, collimator and detector behavior, shot noise affecting the images. This physical model takes into account the full energy dimension of the scattered photons, and can construct a series of images indexed in energy. It is based on the last generation of SPECT (single photon emission computed tomography) medical gamma cameras which operate in list mode (both the coordinates and the energy of the (gamma) photons are measured), in the energy range [50 - 500 keV]. In the model proposed, two hypotheses are made: (1) only the first order scattering is considered, (2) the diffusing medium is homogeneous. This model is computationally light and could lead to the development of new enhancing techniques using the energy dimension.
IEEE/SP 13th Workshop on Statistical Signal Processing, 2005, 2005
The systems commonly used to detect photons and estimate their energies are usually irrelevant fo... more The systems commonly used to detect photons and estimate their energies are usually irrelevant for high flux. Hidden Markov model and jump Markov linear systems (JMLS) provide a framework which allows us to get an optimal estimate for stochastic processes, whose occurrences are randomly distributed according to time of detection, length and magnitude. It is perfectly adapted to the spectrometry
IEEE Signal Processing Letters, 2001
AIP Conference Proceedings, 2001
In scintigraphic imaging, among all undesirable effects photon scattering is undoubtodly the bigg... more In scintigraphic imaging, among all undesirable effects photon scattering is undoubtodly the biggest problem. This leads to an apparent spreading of the gamma ray sources over large areas. This drawback is especially penalizing when detecting small details. Moreover, to this adds the Poisson noise of the photon emission process. Existing restoration methods [1] consist in eliminating as much as possible the diffused photons and only keeping the primary (non-scattered) ones. They lead to the loss of a great number of photons that is equivalent to reduce the signal-to-noise ratio. This situation does not favor the medical applications where the detectability of small lesions is indispensable. We propose a restoration method taking into account the diffused photons. As photon scattering is closely linked to his energy loss, we had previously constructed an image formation model, which creates a series of images indexed in energy [2]. Then, a corresponding method of restoration from the multi-energy image sequence is proposed in this work. Such a restoration has the advantage of exploiting the energy information for improving the image quality. But, in this procedure it is necessary to resolve in parallel two problems: data fusion of all energy channels and restoration in a joint spatial-energy space. We choose here the Bayesian approach because it provides a coherent framework for this task [3]. In our Bayesian estimator, the data fusion is performed in the likelihood term. As photon emission in each energy channel is an independent Poisson process, the global likelihood is the product of the component likelihoods. The difference of the mono-energy likelihood taking account of the only Poisson noise, the multi-energy likelihood contains, in addition, the scattering effect. This allows that the prior can be chosen as a good compromise between the computation time and functional complexity. This is important in the multidimensional case where the computation is heavy. In our work, a convex edge-preserving prior is used [4] that leads to an efficient computation. Application of the proposed restoration method to a multi-energy image series has given encouraging results: the evident improvement of image quality compared to the mono-energy restoration and thus the good recovery of the smallest discontinuities which correspond to the most suspicious diseased areas in an organ. The introduction of multi-energy information opens a new interesting track to scintigraphic image processing.
HAL (Le Centre pour la Communication Scientifique Directe), 1999
Ce travail propose une nouvelle approche de restauration 2D des images scintigraphiques. L'or... more Ce travail propose une nouvelle approche de restauration 2D des images scintigraphiques. L'originalite repose sur la prise en compte de l'energie des photons detectes, grâce a un modele physique realiste et rapide du processus de formation des images. Le modele fournit non pas une, mais une serie d'images indexee en energie, simulant et exploitant les cameras fonctionnant en mode liste energetique. Le modele incorpore la diffusion Compton d'ordre 1, l'attenuation inhomogene par les tissus, et les effets engendres par la camera. Il permet une correction conjointe de ces effets, sans chercher a " eliminer " mais plutot a exploiter les photons diffuses. Le modele est valide par des simulations de Monte Carlo d'ordre 1 et par des mesures effectuees en environnement hospitalier. Son domaine energetique de validite est precise. Cette approche multidimensionnelle ouvre la voie a de nouvelles approches de restauration. Dans un premier temps, une analyse en ...
A method for processing a noisy digital time signal yk of digital pitch k, corresponding to an in... more A method for processing a noisy digital time signal yk of digital pitch k, corresponding to an initial analogue signal st after being conditioned by a conditioning chain. The initial analogue signal st includes at least one pulse representing information concerning at least one radiation from a radiation source, the radiation and the pulse having an energy distribution. The method includes the determination of a non-noisy digital estimation signal sk from the noisy time signal yk by using a state model representing the conditioning imposed by the conditioning chain and in that the state model includes a Markovian variable rk to be estimated whereof at least two values are associated with physical characteristics of at least two typical pulses constituting a possible representation, at least approximate, of the pulse in the signal st.
Medical Imaging 1999: Physics of Medical Imaging, 1999
ABSTRACT This paper proposes a realistic joint energy-spatial description of the various phenomen... more ABSTRACT This paper proposes a realistic joint energy-spatial description of the various phenomena encountered in low energy gamma imagery: Compton diffusion, photoelectric absorption, collimator and detector behavior, shot noise affecting the images. This physical model takes into account the full energy dimension of the scattered photons, and can construct a series of images indexed in energy. It is based on the last generation of SPECT (single photon emission computed tomography) medical gamma cameras which operate in list mode (both the coordinates and the energy of the (gamma) photons are measured), in the energy range [50 - 500 keV]. In the model proposed, two hypotheses are made: (1) only the first order scattering is considered, (2) the diffusing medium is homogeneous. This model is computationally light and could lead to the development of new enhancing techniques using the energy dimension.
IEEE/SP 13th Workshop on Statistical Signal Processing, 2005, 2005
The systems commonly used to detect photons and estimate their energies are usually irrelevant fo... more The systems commonly used to detect photons and estimate their energies are usually irrelevant for high flux. Hidden Markov model and jump Markov linear systems (JMLS) provide a framework which allows us to get an optimal estimate for stochastic processes, whose occurrences are randomly distributed according to time of detection, length and magnitude. It is perfectly adapted to the spectrometry
IEEE Signal Processing Letters, 2001
AIP Conference Proceedings, 2001
In scintigraphic imaging, among all undesirable effects photon scattering is undoubtodly the bigg... more In scintigraphic imaging, among all undesirable effects photon scattering is undoubtodly the biggest problem. This leads to an apparent spreading of the gamma ray sources over large areas. This drawback is especially penalizing when detecting small details. Moreover, to this adds the Poisson noise of the photon emission process. Existing restoration methods [1] consist in eliminating as much as possible the diffused photons and only keeping the primary (non-scattered) ones. They lead to the loss of a great number of photons that is equivalent to reduce the signal-to-noise ratio. This situation does not favor the medical applications where the detectability of small lesions is indispensable. We propose a restoration method taking into account the diffused photons. As photon scattering is closely linked to his energy loss, we had previously constructed an image formation model, which creates a series of images indexed in energy [2]. Then, a corresponding method of restoration from the multi-energy image sequence is proposed in this work. Such a restoration has the advantage of exploiting the energy information for improving the image quality. But, in this procedure it is necessary to resolve in parallel two problems: data fusion of all energy channels and restoration in a joint spatial-energy space. We choose here the Bayesian approach because it provides a coherent framework for this task [3]. In our Bayesian estimator, the data fusion is performed in the likelihood term. As photon emission in each energy channel is an independent Poisson process, the global likelihood is the product of the component likelihoods. The difference of the mono-energy likelihood taking account of the only Poisson noise, the multi-energy likelihood contains, in addition, the scattering effect. This allows that the prior can be chosen as a good compromise between the computation time and functional complexity. This is important in the multidimensional case where the computation is heavy. In our work, a convex edge-preserving prior is used [4] that leads to an efficient computation. Application of the proposed restoration method to a multi-energy image series has given encouraging results: the evident improvement of image quality compared to the mono-energy restoration and thus the good recovery of the smallest discontinuities which correspond to the most suspicious diseased areas in an organ. The introduction of multi-energy information opens a new interesting track to scintigraphic image processing.