A Multimodal Ensemble Deep Learning Model for Functional Outcome Prognosis of Stroke Patients
Journal of Stroke · 31.05.2024 · 26 (2) · 312-320
DOI: 10.5853/jos.2023.03426 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Stroke, erschienen am 31.05.2024. Verfasst von Hye-Soo Jung und 20 weiteren Autor:innen. Beteiligte Einrichtung: Department of Neurology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea. Die Arbeit wurde bisher 24-mal zitiert (Crossref) und nennt 42 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Korea Health Industry Development Institute, Ministry of Health and Welfare, National IT Industry Promotion Agency.
Themen: Schlaganfall
Abstract (Originalsprache)
Background and Purpose The accurate prediction of functional outcomes in patients with acute ischemic stroke (AIS) is crucial for informed clinical decision-making and optimal resource utilization. As such, this study aimed to construct an ensemble deep learning model that integrates multimodal imaging and clinical data to predict the 90-day functional outcomes after AIS.
Methods
We used data from the Korean Stroke Neuroimaging Initiative database, a prospective multicenter stroke registry to construct an ensemble model integrated individual 3D convolutional neural networks for diffusion-weighted imaging and fluid-attenuated inversion recovery (FLAIR), along with a deep neural network for clinical data, to predict 90-day functional independence after AIS using a modified Rankin Scale (mRS) of 3–6. To evaluate the performance of the ensemble model, we compared the area under the curve (AUC) of the proposed method with that of individual models trained on each modality to identify patients with AIS with an mRS score of 3–6.
Results
Of the 2,606 patients with AIS, 993 (38.1%) achieved an mRS score of 3–6 at 90 days post-stroke. Our model achieved AUC values of 0.830 (standard cross-validation [CV]) and 0.779 (time-based CV), which significantly outperformed the other models relying on single modalities: b-value of 1,000 s/mm2 (P
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Neurology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea.
- Department of Neurology, Kyung Hee University Medical Center, Seoul, Korea.
- Department of Neurology, Pusan National University Hospital, Busan, Korea.
- Department of Neurology, Yeungnam University Medical Center, Daegu, Korea.
- Department of Neurology, Dong-A University Hospital, Busan, Korea.
- Department of Neurology, Chonnam National University Hospital, Gwangju, Korea.
Förderung
- Korea Health Industry Development Institute
- Ministry of Health and Welfare
- National IT Industry Promotion Agency
- Ministry of Science and ICT
Zitieren
APA
Jung, H. S., Lee, E. J., Chang, D. I., Cho, H. J., Lee, J., Cha, J. K., Park, M. S., Yu, K. H., Jung, J. M., Ahn, S. H., Kim, D. E., Lee, J. H., Hong, K. S., Sohn, S. I., Park, K. P., Kwon, S. U., Kim, J. S., Chang, J. Y., Kim, B. J., …, & KOSNI Investigators (2024). A Multimodal Ensemble Deep Learning Model for Functional Outcome Prognosis of Stroke Patients. Journal of Stroke, 26(2), 312-320. https://doi.org/10.5853/jos.2023.03426
Vancouver
Jung HS, Lee EJ, Chang DI, Cho HJ, Lee J, Cha JK, et al. A Multimodal Ensemble Deep Learning Model for Functional Outcome Prognosis of Stroke Patients. Journal of Stroke. 2024;26(2):312-320. doi:10.5853/jos.2023.03426
BibTeX
@article{Jung2024,
author = {Jung, Hye-Soo and Lee, Eun-Jae and Chang, Dae-Il and Cho, Han Jin and Lee, Jun and Cha, Jae-Kwan and Park, Man-Seok and Yu, Kyung Ho and Jung, Jin-Man and Ahn, Seong Hwan and Kim, Dong-Eog and Lee, Ju Hun and Hong, Keun-Sik and Sohn, Sung-Il and Park, Kyung-Pil and Kwon, Sun U and Kim, Jong S and Chang, Jun Young and Kim, Bum Joon and Kang, Dong-Wha and KOSNI Investigators, KOSNI Investigators},
title = {A Multimodal Ensemble Deep Learning Model for Functional Outcome Prognosis of Stroke Patients},
journal = {Journal of Stroke},
year = {2024},
volume = {26},
number = {2},
pages = {312--320},
doi = {10.5853/jos.2023.03426}
}
EUVIMED