TY - JOUR

T1 - Voxelisation in the 3-D Fly Algorithm for PET

AU - Abbood, Zainab

AU - Lavauzelle, Julien

AU - Lutton, Evelyne

AU - Rocchisani, Jean-Marie

AU - Louchet, Jean

AU - Vidal, Franck

PY - 2017/10

Y1 - 2017/10

N2 - The Fly Algorithm was initially developed for 3-D robot vision applications. It consists in solving the inverse problem of shape reconstruction from projections by evolving a population of 3-D points in space (the ‘flies’), using an evolutionary optimisation strategy. Here, in its version dedicated to tomographic reconstruction in medical imaging, the flies are mimicking radioactive photon sources. Evolution is controlled using a fitness function based on the discrepancy of the projections simulated by the flies with the actual pattern received by the sensors. The reconstructed radioactive concentration is derived from the population of flies, i.e. a collection of points in the 3-D Euclidean space, after convergence. ‘Good’ flies were previously binned into voxels. In this paper, we study which flies to include in the final solution and how this information can be sampled to provide more accurate datasets in a reduced computation time. We investigate the use of density fields, based on Metaballs and on Gaussian functions respectively, to obtain a realistic output. The spread of each Gaussian kernel is modulated in function of the corresponding fly fitness. The resulting volumes are compared with previous work in terms of normalised-cross correlation. In our test-cases, data fidelity increases by more than 10% when density fields are used instead of binning. Our method also provides reconstructions comparable to those obtained using well-established techniques used in medicine (filtered back-projection and ordered subset expectation-maximisation).

AB - The Fly Algorithm was initially developed for 3-D robot vision applications. It consists in solving the inverse problem of shape reconstruction from projections by evolving a population of 3-D points in space (the ‘flies’), using an evolutionary optimisation strategy. Here, in its version dedicated to tomographic reconstruction in medical imaging, the flies are mimicking radioactive photon sources. Evolution is controlled using a fitness function based on the discrepancy of the projections simulated by the flies with the actual pattern received by the sensors. The reconstructed radioactive concentration is derived from the population of flies, i.e. a collection of points in the 3-D Euclidean space, after convergence. ‘Good’ flies were previously binned into voxels. In this paper, we study which flies to include in the final solution and how this information can be sampled to provide more accurate datasets in a reduced computation time. We investigate the use of density fields, based on Metaballs and on Gaussian functions respectively, to obtain a realistic output. The spread of each Gaussian kernel is modulated in function of the corresponding fly fitness. The resulting volumes are compared with previous work in terms of normalised-cross correlation. In our test-cases, data fidelity increases by more than 10% when density fields are used instead of binning. Our method also provides reconstructions comparable to those obtained using well-established techniques used in medicine (filtered back-projection and ordered subset expectation-maximisation).

KW - Fly algorithm

KW - Evolutionary computation

KW - Tomography

KW - Reconstruction algorithms

KW - Iterative algorithms

KW - Inverse problems

KW - Iterative reconstruction

KW - Co-operative co-evolution

U2 - 10.1016/j.swevo.2017.04.001

DO - 10.1016/j.swevo.2017.04.001

M3 - Article

VL - 36

SP - 91

EP - 105

JO - Swarm and Evolutionary Computation

JF - Swarm and Evolutionary Computation

SN - 2210-6502

ER -