TY - JOUR
T1 - OTONet
T2 - Deep Neural Network for Precise Otoscopy Image Classification
AU - Rao, Divya
AU - Singh, Rohit
AU - Kamath, Sudiksha
AU - Pendekanti, Sanjeev
AU - Pai, Divya
AU - Kolekar, Sucheta
AU - Holla, Raviraj
AU - Pathan, Sameena
N1 - Publisher Copyright:
© 2013 IEEE.
PY - 2024
Y1 - 2024
N2 - Otoscopy is a diagnostic procedure to visualize the external ear canal and eardrum, facilitating the detection of various ear pathologies and conditions. Timely otoscopy image classification offers significant advantages, including early detection, reduced patient anxiety, and personalized treatment plans. This paper introduces a novel OTONet framework specifically tailored for otoscopy image classification. It leverages octave 3D convolution and a combination of feature and region-focus modules to create an accurate and robust classification system capable of distinguishing between various otoscopic conditions. This architecture is designed to efficiently capture and process the spatial and feature information present in otoscopy images. Using a public otoscopy dataset, OTONet has reached a classification accuracy of 99.3% and an F1 score of 99.4% across 11 classes of ear conditions. A comparative analysis demonstrates that OTONet surpasses other established machine learning models, including ResNet50, ResNet50v2, VGG16, Dense-Net169, and ConvNeXtTiny, across various evaluation metrics. The research's contribution to improved diagnostic accuracy reduced human error, expedited diagnostics, and its potential for telemedicine applications.
AB - Otoscopy is a diagnostic procedure to visualize the external ear canal and eardrum, facilitating the detection of various ear pathologies and conditions. Timely otoscopy image classification offers significant advantages, including early detection, reduced patient anxiety, and personalized treatment plans. This paper introduces a novel OTONet framework specifically tailored for otoscopy image classification. It leverages octave 3D convolution and a combination of feature and region-focus modules to create an accurate and robust classification system capable of distinguishing between various otoscopic conditions. This architecture is designed to efficiently capture and process the spatial and feature information present in otoscopy images. Using a public otoscopy dataset, OTONet has reached a classification accuracy of 99.3% and an F1 score of 99.4% across 11 classes of ear conditions. A comparative analysis demonstrates that OTONet surpasses other established machine learning models, including ResNet50, ResNet50v2, VGG16, Dense-Net169, and ConvNeXtTiny, across various evaluation metrics. The research's contribution to improved diagnostic accuracy reduced human error, expedited diagnostics, and its potential for telemedicine applications.
UR - http://www.scopus.com/inward/record.url?scp=85182376743&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85182376743&partnerID=8YFLogxK
U2 - 10.1109/ACCESS.2024.3351668
DO - 10.1109/ACCESS.2024.3351668
M3 - Article
AN - SCOPUS:85182376743
SN - 2169-3536
VL - 12
SP - 7734
EP - 7746
JO - IEEE Access
JF - IEEE Access
ER -