GlottisNetV2: Temporal Glottal Midline Detection using Deep Convolutional Neural Networks

Kruse E, Döllinger M, Schützenberger A, Kist A (2023)

Publication Type: Journal article

Publication year: 2023

Journal

IEEE Journal of Translational Engineering in Health and Medicine Institute of Electrical and Electronics Engineers (IEEE)

Pages Range: 1-1

DOI: 10.1109/JTEHM.2023.3237859

Abstract

Objective: High-speed videoendoscopy is a major tool for quantitative laryngology. Glottis segmentation and glottal midline detection are crucial for computing vocal fold-specific, quantitative parameters. However, fully automated solutions show limited clinical applicability. Especially unbiased glottal midline detection remains a challenging problem. Methods and procedures: We developed a multitask deep neural network for glottis segmentation and glottal midline detection. We used techniques from pose estimation to estimate the anterior and posterior points in endoscopy images. Neural networks were set up in TensorFlow/Keras and trained and evaluated with the BAGLS dataset. Results: We found that a dual decoder deep neural network termed GlottisNetV2 outperforms the previously proposed GlottisNet in terms of MAPE on the test dataset (1.85% to 6.3%) while converging faster. Using various hyperparameter tunings, we allow fast and directed training. Using temporal variant data on an additional data set designed for this task, we can improve the median prediction accuracy from 2.1% to 1.76% when using 12 consecutive frames and additional temporal filtering. Conclusion: We found that temporal glottal midline detection using a dual decoder architecture together with keypoint estimation allows accurate midline prediction. Clinical impact: We show that our proposed architecture allows stable and reliable glottal midline predictions ready for clinical use and analysis of symmetry measures.

Authors with CRIS profile

Michael Döllinger Professur für Computational Medicine (DFG-Forschungsprofessur) Anne Schützenberger Department of Otorhinolaryngology – Head and Neck Surgery Andreas Kist Juniorprofessur für Artificial Intelligence in Communication Disorders

Related research project(s)

Objective analysis of functional based hoarseness by clinical high-speed endoscopy (DO 1247/8-1/2) Sept. 28, 2016 - Feb. 15, 2024

How to cite

APA:

Kruse, E., Döllinger, M., Schützenberger, A., & Kist, A. (2023). GlottisNetV2: Temporal Glottal Midline Detection using Deep Convolutional Neural Networks. IEEE Journal of Translational Engineering in Health and Medicine, 1-1. https://doi.org/10.1109/JTEHM.2023.3237859

MLA:

Kruse, Elina, et al. "GlottisNetV2: Temporal Glottal Midline Detection using Deep Convolutional Neural Networks." IEEE Journal of Translational Engineering in Health and Medicine (2023): 1-1.

BibTeX: Download