In this work-in-progress paper, we present preliminary results of a large pilot study for implementing a novel machine learning approach presented at HICSS 2019 [1] and ICIS 2019 [2] in a German hospital to detect epileptic episodes in EEG data. While the algorithm achieved a balanced accuracy of 75.6% on real clinical data we could gain valuable experience regarding the implementation barriers of machine learning algorithms in practice, which is discussed in this paper. These lessons learned have practical implications for future work.
CITATION STYLE
Rieg, T., Frick, J., & Buettner, R. (2020). Machine Learning-Based Diagnosis of Epilepsy in Clinical Routine: Lessons Learned from a Retrospective Pilot Study. In Lecture Notes in Information Systems and Organisation (Vol. 43, pp. 276–283). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-60073-0_32
Mendeley helps you to discover research relevant for your work.