Tympanic membrane segmentation in otoscopic images based on fully convolutional network with active contour loss
- PDF / 927,413 Bytes
- 9 Pages / 595.276 x 790.866 pts Page_size
- 31 Downloads / 191 Views
ORIGINAL PAPER
Tympanic membrane segmentation in otoscopic images based on fully convolutional network with active contour loss Van-Truong Pham1,2 · Thi-Thao Tran1,2 · Pa-Chun Wang3,4 · Men-Tzung Lo2 Received: 14 January 2020 / Revised: 22 July 2020 / Accepted: 25 August 2020 © Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract This paper presents a method to automatically segment tympanic membranes (TMs) from video-otoscopic images based on the deep learning approach. The paper introduces a hybrid loss function combining the Dice loss and active contour loss to the fully convolutional network. By this way, the proposed model takes into account the Dice similarity and the desired boundary contour information including the contour length as well as regions inside and outside the contour during learning. The proposed loss function is then applied to the fully convolutional network for tympanic membrane segmentation. We evaluate the proposed approach on TMs data set which includes 1139 otoscopic images from patients diagnosed with and without otitis media. Experimental results show that the proposed deep learning model achieves an average Dice similarity coefficient of 0.895, a mean Hausdorff distance of 19.189, and average perpendicular distance of 6.429, that outperforms other state-of-the-art methods. Keywords Image segmentation · Active contour model · Deep learning · Otitis media · Tympanic membranes segmentation
1 Introduction Otitis media is a common ear infection among pediatric population, and might lead to severe life-threatening otological or intracranial complications [1]. Otitis media (OM) is defined as any inflammation of the middle ear and can be classified clinically as acute otitis media (AOM), otitis media with effusion (OME) and chronic otitis media (COM). AOM is caused by active bacterial infection leading to congestion or purulence in the middle ear. OME is a consequence of AOM characterized by accumulation of fluid in the middle ear space after acute stage. COM is the chronic stage of
B B
Thi-Thao Tran [email protected] Men-Tzung Lo [email protected]
1
School of Electrical Engineering, Hanoi University of Science and Technology, Hanoi, Vietnam
2
Department of Biomedical Sciences and Engineering, National Central University, Taoyüan, Taiwan
3
Department of Otolaryngology, Cathay General Hospital, Taipei, Taiwan
4
School of Medicine, Fu Jen Catholic University, Taipei, Taiwan
AOM that involves perforation, retraction, or collapse of tympanic membrane following active bacterial infection within the middle ear space [2]. OM can incur significant economic impacts to the society, especially in countries with a vast territory where access to appropriate medical care is not readily available [3]. In fact, OM diseases are related to abnormality of the tympanic membrane (TM) [4, 5]. Establishment of OM diagnosis relies on the identification of morphological or color changes on TM. The tympanic membrane, also known as the eardrum, is a thin membrane dividing the ear canal
Data Loading...
