Publication Details
Evolutionary Design of Reduced Precision Levodopa-Induced Dyskinesia Classifiers
Drahošová Michaela, Ing., Ph.D. (DCSY)
Sekanina Lukáš, prof. Ing., Ph.D. (DCSY)
SMITH, S.
ALTY, J.
Cartesian genetic programming, Coevolution, Adaptive size fitness predictors,
Energy-efficient, Hardware-oriented, Fixed-point arithmetic, Levodopa-induced
dyskinesia, Parkinsons disease
Parkinson's disease is one of the most common neurological conditions whose
symptoms are usually treated with a drug containing levodopa. To minimise
levodopa side effects, i.e. levodopa-induced dyskinesia (LID), it is necessary to
correctly manage levodopa dosage. This article covers an application of cartesian
genetic programming (CGP) to assess LID based on time series collected using
accelerators attached to the patient's body. Evolutionary design of reduced
precision classifiers of LID is investigated in order to find
a hardware-efficient classifier together with classification accuracy as close as
possible to a baseline software implementation. CGP equipped with the coevolution
of adaptive size fitness predictors (coASFP) is used to design LID-classifiers
working with fixed-point arithmetics with reduced precision, which is suitable
for implementation in application-specific integrated circuits. In this
particular task, we achieved a significant evolutionary design computational cost
reduction in comparison with the original CGP. Moreover, coASFP effectively
prevented overfitting in this task. Experiments with reduced precision
LID-classifier design show that evolved classifiers working with 8-bit unsigned
integer data representation, together with the input data scaling using the
logical right shift, not only significantly outperformed hardware characteristics
of all other investigated solutions but also achieved a better classifier
accuracy in comparison with classifiers working with the floating-point numbers.
@inproceedings{BUT177631,
author="HURTA, M. and DRAHOŠOVÁ, M. and SEKANINA, L. and SMITH, S. and ALTY, J.",
title="Evolutionary Design of Reduced Precision Levodopa-Induced Dyskinesia Classifiers",
booktitle="Genetic Programming, 25th European Conference, EuroGP 2022",
year="2022",
series="Lecture Notes in Computer Science",
volume="13223",
pages="85--101",
publisher="Springer Nature Switzerland AG",
address="Madrid",
doi="10.1007/978-3-031-02056-8\{_}6",
isbn="978-3-031-02055-1",
url="https://link.springer.com/chapter/10.1007/978-3-031-02056-8_6"
}