Deep Learning Techniques for Ear Diseases Based on Segmentation of the Normal Tympanic Membrane
1
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
Objectives. Otitis media is a common infection worldwide. Owing to the limited number of ear specialists and rapid devel-opment of telemedicine, several trials have been conducted to develop novel diagnostic strategies to improve the di-agnostic accuracy and screening of patients with otologic diseases based on abnormal otoscopic findings. Althoughthese strategies have demonstrated high diagnostic accuracy for the tympanic membrane (TM), the insufficient ex-plainability of these techniques limits their deployment in clinical practice. Methods. We used a deep convolutional neural network (CNN) model based on the segmentation of a normal TM into fivesubstructures (malleus, umbo, cone of light, pars flaccida, and annulus) to identify abnormalities in otoscopic ear im-ages. The mask R-CNN algorithm learned the labeled images. Subsequently, we evaluated the diagnostic performanceof combinations of the five substructures using a three-layer fully connected neural network to determine whetherear disease was present. Results. We obtained the receiver operating characteristic (ROC) curve of the optimal conditions for the presence or ab-sence of eardrum diseases according to each substructure separately or combinations of substructures. The highestarea under the curve (0.911) was found for a combination of the malleus, cone of light, and umbo, compared withthe corresponding areas under the curve of 0.737–0.873 for each substructure. Thus, an algorithm using these fiveimportant normal anatomical structures could prove to be explainable and effective in screening abnormal TMs. Conclusion. This automated algorithm can improve diagnostic accuracy by discriminating between normal and abnormalTMs and can facilitate appropriate and timely referral consultations to improve patients’ quality of life in the contextof primary care.
Abstract
Objectives. Otitis media is a common infection worldwide. Owing to the limited number of ear specialists and rapid devel-opment of telemedicine, several trials have been conducted to develop novel diagnostic strategies to improve the di-agnostic accuracy and screening of patients with otologic diseases based on abnormal otoscopic findings. Althoughthese strategies have demonstrated high diagnostic accuracy for the tympanic membrane (TM), the insufficient ex-plainability of these techniques limits their deployment in clinical practice. Methods. We used a deep convolutional neural network (CNN) model based on the segmentation of a normal TM into fivesubstructures (malleus, umbo, cone of light, pars flaccida, and annulus) to identify abnormalities in otoscopic ear im-ages. The mask R-CNN algorithm learned the labeled images. Subsequently, we evaluated the diagnostic performanceof combinations of the five substructures using a three-layer fully connected neural network to determine whetherear disease was present. Results. We obtained the receiver operating characteristic (ROC) curve of the optimal conditions for the presence or ab-sence of eardrum diseases according to each substructure separately or combinations of substructures. The highestarea under the curve (0.911) was found for a combination of the malleus, cone of light, and umbo, compared withthe corresponding areas under the curve of 0.737–0.873 for each substructure. Thus, an algorithm using these fiveimportant normal anatomical structures could prove to be explainable and effective in screening abnormal TMs. Conclusion. This automated algorithm can improve diagnostic accuracy by discriminating between normal and abnormalTMs and can facilitate appropriate and timely referral consultations to improve patients’ quality of life in the contextof primary care.
참고문헌 (22)
- Joe H, "A newly designed tympanostomy stent with TiO2 coating to reduce Pseudomonas aeruginosa biofilm formation", J Biomater Appl, 33(4), pp.599~605 (2018)
- Lee SH, "Effects of reactive oxygen species generation induced by Wonju City particulate matter on mitochondrial dysfunction in human middle ear cell", Environ Sci Pollut Res Int, 28(35), pp.49244~49257 (2021)
- Demant MN, "Smartphone otoscopy by non-specialist health workers in rural Greenland : a cross-sectional study", Int J Pediatr Otorhinolaryngol, 126, pp.109628~ (2019)
- Cha D, "Automated diagnosis of ear disease using ensemble deep learning with a big otoendoscopy image database", EBioMedicine, 45, pp.606~614 (2019)
- Zeng X, "Efficient and accurate identification of ear diseases using an ensemble deep learning model", Sci Rep, 11(1), pp.10839~ (2021)
- Singh A, "Explainable deep learning models in medical image analysis", J Imaging, 6(6), pp.52~ (2020)
- Liu X, "A review of deep-learning-based medical image segmentation methods", Sustainability, 13(3), pp.1224~ (2021)
- Rosenfeld RM, "clinical practice guideline : otitis media with effusion(update)", Otolaryngol Head Neck Surg, 154(1 Suppl), pp.S1~41 (2016)
- Sanna M, "Color atlas of endo-otoscopy", Thieme; (2017)
- Russell BC, "LabelMe : a database and web-based tool for image annotation", Int J Comput Vis, 77(1), pp.157~173 (2008)
- He K, "Mask R-CNN", International Conference on Computer Vision, pp.2980~2988 (2017)
- Peng J, "Medical image segmentation with limited supervision : a review of deep network models", IEEE Access, 9, pp.36827~36851 (2021)
- Aggarwal R, "Diagnostic accuracy of deep learning in medical imaging : a systematic review and meta-analysis", NPJ Digit Med, 4(1), pp.65~ (2021)
- Wang G, "Automatic brain tumor segmentation using cascaded anisotropic convolutional neural networks", Brainlesion: glioma, multiple sclerosis, stroke and traumatic brain injuries. Proceedings of the Third International Workshop BrainLes, pp.10670~ (2017), Springer
- Liu Y, "Automatic segmentation of cervical nuclei based on deep learning and a conditional random field", IEEE Access, 6, pp.53709~53721 (2018)
- Zhao C, "Lung nodule detection via 3D U-Net and contextual convolutional neural network", 2018 International Conference on Networking and Network Applications, pp.356~361 (2018)
- Mulay S, "Liver segmentation from multimodal images using HED-Mask R-CNN", Multiscale multimodal medical imaging. Proceedings of the First International Workshop MMMI 2019, pp.68~75 (2019)
- Shu JH, "An improved mask R-CNN model for multiorgan segmentation", Math Probl Eng, 2020, pp.8351725~ (2020)
- Prajapati SA, "Classification of dental diseases using CNN and transfer learning", 5th International Symposium on Computational and Business Intelligence (ISCBI), pp.70~74 (2017)
- Anantharaman R, "Utilizing Mask R-CNN for detection and segmentation of oral diseases", 2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp.2197~2204 (2018)
- Myburgh HC, "Otitis media diagnosis for developing countries using tympanic membrane image-analysis", EBioMedicine, 5, pp.156~160 (2016)
- Pichichero ME, "Assessing diagnostic accuracy and tympanocentesis skills in the management of otitis media", Arch Pediatr Adolesc Med, 155(10), pp.1137~1142 (2001)