Investigating the use of a two-stage attention-aware convolutional neural network for the automated diagnosis of otitis media from tympanic membrane images: a prediction model development and validation study
MetadataShow full item record
Objectives This study investigated the usefulness and performance of a two-stage attention-aware convolutional neural network (CNN) for the automated diagnosis of otitis media from tympanic membrane (TM) images. Design A classification model development and validation study in ears with otitis media based on otoscopic TM images. Two commonly used CNNs were trained and evaluated on the dataset. On the basis of a Class Activation Map (CAM), a two-stage classification pipeline was developed to improve accuracy and reliability, and simulate an expert reading the TM images. Setting and participants This is a retrospective study using otoendoscopic images obtained from the Department of Otorhinolaryngology in China. A dataset was generated with 6066 otoscopic images from 2022 participants comprising four kinds of TM images, that is, normal eardrum, otitis media with effusion (OME) and two stages of chronic suppurative otitis media (CSOM). Results The proposed method achieved an overall accuracy of 93.4% using ResNet50 as the backbone network in a threefold cross-validation. The F1 Score of classification for normal images was 94.3%, and 96.8% for OME. There was a small difference between the active and inactive status of CSOM, achieving 91.7% and 82.4% F1 scores, respectively. The results demonstrate a classification performance equivalent to the diagnosis level of an associate professor in otolaryngology. Conclusions CNNs provide a useful and effective tool for the automated classification of TM images. In addition, having a weakly supervised method such as CAM can help the network focus on discriminative parts of the image and improve performance with a relatively small database. This two-stage method is beneficial to improve the accuracy of diagnosis of otitis media for junior otolaryngologists and physicians in other disciplines.
Cai, Y., Yu, J.G., Chen, Y., Liu, C., Xiao, L., Grais, E.M., Zhao, F., Lan, L., Zeng, S., Zeng, J. and Wu, M. (2021) 'Investigating the use of a two-stage attention-aware convolutional neural network for the automated diagnosis of otitis media from tympanic membrane images: a prediction model development and validation study', BMJ open, 11(1), p.e041139.
Article published in BMJ Open Ear, nose and throat/otolaryngology available open access at http://dx.doi.org/10.1136/bmjopen-2020-041139
This work was supported by the Key R&D Programme of Guangdong Province, China (Grant No. 2018B030339001), medical artificial intelligence project of Sun Yat-Sen Memorial Hospital (YXYGZN201904) and the National Natural Science Foundation of China (Grant No. 81570935)
Showing items related by title, author, subject and abstract.
Yang, Jiaji; Chew, Esyin (Springer, 2021-01-25)With the rapid development of traditional industries, intelligent robots have been widely used in the hospitality industry. Although the development of intelligent robots faces a positive trend and a good market in the ...
An investigation into the effects that the social media platforms Facebook and Pinterest have on the planning stage of weddings Cradduck, Paige (Cardiff Metropolitan University, 2018-04)This dissertation attempts to identify to what extent social media is impacting how people plan their weddings. Social media has become more and more popular over the years, therefore this study researches how the growth ...
Bohadana, Ruti Thalia (Cardiff Metropolitan University, 2017-06-01)Background: Obesity is at epidemic level in the UK with 25% of adults estimated to be obese in 2013 (HSE, 2014). Many people look to the media for health and nutrition information (Hill, 2006). It has been found that the ...