Classification of Wideband Tympanometry by Deep Transfer Learning With Data Augmentation for Automatic Diagnosis of Otosclerosis
Leixin Nie,Chao Li,Franck Marzani,Haibin Wang,Francois Thibouw,Alexis Bozorg Grayeli,Leixin Nie,Chao Li,Franck Marzani,Haibin Wang,Alexis Bozorg Grayeli
DOI: https://doi.org/10.1109/jbhi.2021.3093007
IF: 7.7
2022-02-01
IEEE Journal of Biomedical and Health Informatics
Abstract:Otosclerosis is a common disease of the middle ear leading to stapedial fixation. Its rapid and non-invasive diagnosis could be achieved through wideband tympanometry (WBT), but the interpretation of the raw data provided by this tool is complex and time-consuming. Convolutional neural networks (CNN) could potentially be applied to this situation to help the clinicians categorize WBT data. A dataset containing 135 samples from 80 patients with otosclerosis and 55 controls was obtained. We designed a lightweight CNN to categorize samples into the otosclerosis and control. Receiver operating characteristic (ROC) analysis showed an area under the curve (AUC) of 0.95 ±0.011, and the F1-score was 0.89 ±0.031 ( r=10). The performance was further improved by data augmentation schemes and transfer learning strategies (AUC: 0.97 ±0.010, F1-score: 0.94 ±0.016, , ANOVA). Finally, the most relevant diagnostic features employed by the CNN were assessed via the activation pattern heatmaps. These results are crucial for the visual interpretation of WBT graphic outputs which clinicians use in routine, and for a better understanding of the WBT signal in relation to the ossicular mechanics.
computer science, interdisciplinary applications,mathematical & computational biology,medical informatics, information systems