Data exploration in evolutionary reconstruction of PET images

Research output: Contribution to journalArticle

Standard Standard

Data exploration in evolutionary reconstruction of PET images. / Gray, Cameron; Al-Maliki, Shatha; Vidal, Franck.

In: Genetic Programming and Evolvable Machines, Vol. 19, No. 3, 09.2018, p. 391-419.

Research output: Contribution to journalArticle

HarvardHarvard

Gray, C, Al-Maliki, S & Vidal, F 2018, 'Data exploration in evolutionary reconstruction of PET images', Genetic Programming and Evolvable Machines, vol. 19, no. 3, pp. 391-419. https://doi.org/10.1007/s10710-018-9330-7

APA

Gray, C., Al-Maliki, S., & Vidal, F. (2018). Data exploration in evolutionary reconstruction of PET images. Genetic Programming and Evolvable Machines, 19(3), 391-419. https://doi.org/10.1007/s10710-018-9330-7

CBE

Gray C, Al-Maliki S, Vidal F. 2018. Data exploration in evolutionary reconstruction of PET images. Genetic Programming and Evolvable Machines. 19(3):391-419. https://doi.org/10.1007/s10710-018-9330-7

MLA

VancouverVancouver

Gray C, Al-Maliki S, Vidal F. Data exploration in evolutionary reconstruction of PET images. Genetic Programming and Evolvable Machines. 2018 Sep;19(3):391-419. https://doi.org/10.1007/s10710-018-9330-7

Author

Gray, Cameron ; Al-Maliki, Shatha ; Vidal, Franck. / Data exploration in evolutionary reconstruction of PET images. In: Genetic Programming and Evolvable Machines. 2018 ; Vol. 19, No. 3. pp. 391-419.

RIS

TY - JOUR

T1 - Data exploration in evolutionary reconstruction of PET images

AU - Gray, Cameron

AU - Al-Maliki, Shatha

AU - Vidal, Franck

PY - 2018/9

Y1 - 2018/9

N2 - This work is based on a cooperation co-evolution algorithm called `Fly Algorithm', which is an evolutionary algorithm (EA) where individuals are called `flies'. It is a specific case of the `Parisian Approach' where the solution of an optimisation problem is a set of individuals (e.g. the whole population) instead of a single individual (the best one) as in typical EAs. The optimisation problem considered here is tomography reconstruction in positron emission tomography (PET). It estimates the concentration of a radioactive substance (called a radiotracer) within the body. Tomography, in this context, is considered as a difficult ill-posed inverse problem. The Fly Algorithm aims at optimising the position of 3-D points that mimic the radiotracer. At the end of the optimisation process, the population of flies is extracted as it corresponds to an estimate of the radioactive substance concentration. During the optimisation loop a lot of data is generated by the algorithm, such as image metrics, duration, and internal states. This data is recorded in a log file that can be post-processed and visualised. We propose using information visualisation and user interaction techniques to explore the algorithm's internal data after reconstruction. Our aim is to better understand what happens during the evolutionary loop. Using an example, we demonstrate that it is possible to interactively discover when an early termination could be triggered. It is implemented in a new stopping criterion, which leads to a 60% reduction of the number of iterations without any loss of accuracy.

AB - This work is based on a cooperation co-evolution algorithm called `Fly Algorithm', which is an evolutionary algorithm (EA) where individuals are called `flies'. It is a specific case of the `Parisian Approach' where the solution of an optimisation problem is a set of individuals (e.g. the whole population) instead of a single individual (the best one) as in typical EAs. The optimisation problem considered here is tomography reconstruction in positron emission tomography (PET). It estimates the concentration of a radioactive substance (called a radiotracer) within the body. Tomography, in this context, is considered as a difficult ill-posed inverse problem. The Fly Algorithm aims at optimising the position of 3-D points that mimic the radiotracer. At the end of the optimisation process, the population of flies is extracted as it corresponds to an estimate of the radioactive substance concentration. During the optimisation loop a lot of data is generated by the algorithm, such as image metrics, duration, and internal states. This data is recorded in a log file that can be post-processed and visualised. We propose using information visualisation and user interaction techniques to explore the algorithm's internal data after reconstruction. Our aim is to better understand what happens during the evolutionary loop. Using an example, we demonstrate that it is possible to interactively discover when an early termination could be triggered. It is implemented in a new stopping criterion, which leads to a 60% reduction of the number of iterations without any loss of accuracy.

U2 - 10.1007/s10710-018-9330-7

DO - 10.1007/s10710-018-9330-7

M3 - Article

VL - 19

SP - 391

EP - 419

JO - Genetic Programming and Evolvable Machines

JF - Genetic Programming and Evolvable Machines

SN - 1389-2576

IS - 3

ER -