Automated classification of osteomeatal complex inflammation on computed tomography using convolutional neural networks

Naweed I. Chowdhury, Timothy Smith, Rakesh K. Chandra, Justin H. Turner

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Background: Convolutional neural networks (CNNs) are advanced artificial intelligence algorithms well suited to image classification tasks with variable features. These have been used to great effect in various real-world applications including handwriting recognition, face detection, image search, and fraud prevention. We sought to retrain a robust CNN with coronal computed tomography (CT) images to classify osteomeatal complex (OMC) occlusion and assess the performance of this technology with rhinologic data. Methods: The Google Inception-V3 CNN trained with 1.28 million images was used as the base model. Preoperative coronal sections through the OMC were obtained from 239 patients enrolled in 2 prospective chronic rhinosinusitis (CRS) outcomes studies, labeled according to OMC status, and mirrored to obtain a set of 956 images. Using this data, the classification layer of Inception-V3 was retrained in Python using a transfer learning method to adapt the CNN to the task of interpreting sinonasal CT images. Results: The retrained neural network achieved 85% classification accuracy for OMC occlusion, with a 95% confidence interval for algorithm accuracy of 78% to 92%. Receiver operating characteristic (ROC) curve analysis on the test set confirmed good classification ability of the CNN with an area under the ROC curve (AUC) of 0.87, significantly different than both random guessing and a dominant classifier that predicts the most common class (p < 0.0001). Conclusion: Current state-of-the-art CNNs may be able to learn clinically relevant information from 2-dimensional sinonasal CT images with minimal supervision. Future work will extend this approach to 3-dimensional images in order to further refine this technology for possible clinical applications.

Original languageEnglish (US)
JournalInternational Forum of Allergy and Rhinology
DOIs
StateAccepted/In press - Jan 1 2018

Fingerprint

Tomography
Inflammation
ROC Curve
Boidae
Handwriting
Technology
Fraud
Aptitude
Artificial Intelligence
Outcome Assessment (Health Care)
Confidence Intervals

Keywords

  • chronic disease
  • convolutional neural network
  • machine learning
  • neural network
  • sinusitis

ASJC Scopus subject areas

  • Immunology and Allergy
  • Otorhinolaryngology

Cite this

Automated classification of osteomeatal complex inflammation on computed tomography using convolutional neural networks. / Chowdhury, Naweed I.; Smith, Timothy; Chandra, Rakesh K.; Turner, Justin H.

In: International Forum of Allergy and Rhinology, 01.01.2018.

Research output: Contribution to journalArticle

@article{c06a3061b43e4d768cbcdc58cc105ab6,
title = "Automated classification of osteomeatal complex inflammation on computed tomography using convolutional neural networks",
abstract = "Background: Convolutional neural networks (CNNs) are advanced artificial intelligence algorithms well suited to image classification tasks with variable features. These have been used to great effect in various real-world applications including handwriting recognition, face detection, image search, and fraud prevention. We sought to retrain a robust CNN with coronal computed tomography (CT) images to classify osteomeatal complex (OMC) occlusion and assess the performance of this technology with rhinologic data. Methods: The Google Inception-V3 CNN trained with 1.28 million images was used as the base model. Preoperative coronal sections through the OMC were obtained from 239 patients enrolled in 2 prospective chronic rhinosinusitis (CRS) outcomes studies, labeled according to OMC status, and mirrored to obtain a set of 956 images. Using this data, the classification layer of Inception-V3 was retrained in Python using a transfer learning method to adapt the CNN to the task of interpreting sinonasal CT images. Results: The retrained neural network achieved 85{\%} classification accuracy for OMC occlusion, with a 95{\%} confidence interval for algorithm accuracy of 78{\%} to 92{\%}. Receiver operating characteristic (ROC) curve analysis on the test set confirmed good classification ability of the CNN with an area under the ROC curve (AUC) of 0.87, significantly different than both random guessing and a dominant classifier that predicts the most common class (p < 0.0001). Conclusion: Current state-of-the-art CNNs may be able to learn clinically relevant information from 2-dimensional sinonasal CT images with minimal supervision. Future work will extend this approach to 3-dimensional images in order to further refine this technology for possible clinical applications.",
keywords = "chronic disease, convolutional neural network, machine learning, neural network, sinusitis",
author = "Chowdhury, {Naweed I.} and Timothy Smith and Chandra, {Rakesh K.} and Turner, {Justin H.}",
year = "2018",
month = "1",
day = "1",
doi = "10.1002/alr.22196",
language = "English (US)",
journal = "International Forum of Allergy and Rhinology",
issn = "2042-6976",
publisher = "Wiley-Blackwell",

}

TY - JOUR

T1 - Automated classification of osteomeatal complex inflammation on computed tomography using convolutional neural networks

AU - Chowdhury, Naweed I.

AU - Smith, Timothy

AU - Chandra, Rakesh K.

AU - Turner, Justin H.

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Background: Convolutional neural networks (CNNs) are advanced artificial intelligence algorithms well suited to image classification tasks with variable features. These have been used to great effect in various real-world applications including handwriting recognition, face detection, image search, and fraud prevention. We sought to retrain a robust CNN with coronal computed tomography (CT) images to classify osteomeatal complex (OMC) occlusion and assess the performance of this technology with rhinologic data. Methods: The Google Inception-V3 CNN trained with 1.28 million images was used as the base model. Preoperative coronal sections through the OMC were obtained from 239 patients enrolled in 2 prospective chronic rhinosinusitis (CRS) outcomes studies, labeled according to OMC status, and mirrored to obtain a set of 956 images. Using this data, the classification layer of Inception-V3 was retrained in Python using a transfer learning method to adapt the CNN to the task of interpreting sinonasal CT images. Results: The retrained neural network achieved 85% classification accuracy for OMC occlusion, with a 95% confidence interval for algorithm accuracy of 78% to 92%. Receiver operating characteristic (ROC) curve analysis on the test set confirmed good classification ability of the CNN with an area under the ROC curve (AUC) of 0.87, significantly different than both random guessing and a dominant classifier that predicts the most common class (p < 0.0001). Conclusion: Current state-of-the-art CNNs may be able to learn clinically relevant information from 2-dimensional sinonasal CT images with minimal supervision. Future work will extend this approach to 3-dimensional images in order to further refine this technology for possible clinical applications.

AB - Background: Convolutional neural networks (CNNs) are advanced artificial intelligence algorithms well suited to image classification tasks with variable features. These have been used to great effect in various real-world applications including handwriting recognition, face detection, image search, and fraud prevention. We sought to retrain a robust CNN with coronal computed tomography (CT) images to classify osteomeatal complex (OMC) occlusion and assess the performance of this technology with rhinologic data. Methods: The Google Inception-V3 CNN trained with 1.28 million images was used as the base model. Preoperative coronal sections through the OMC were obtained from 239 patients enrolled in 2 prospective chronic rhinosinusitis (CRS) outcomes studies, labeled according to OMC status, and mirrored to obtain a set of 956 images. Using this data, the classification layer of Inception-V3 was retrained in Python using a transfer learning method to adapt the CNN to the task of interpreting sinonasal CT images. Results: The retrained neural network achieved 85% classification accuracy for OMC occlusion, with a 95% confidence interval for algorithm accuracy of 78% to 92%. Receiver operating characteristic (ROC) curve analysis on the test set confirmed good classification ability of the CNN with an area under the ROC curve (AUC) of 0.87, significantly different than both random guessing and a dominant classifier that predicts the most common class (p < 0.0001). Conclusion: Current state-of-the-art CNNs may be able to learn clinically relevant information from 2-dimensional sinonasal CT images with minimal supervision. Future work will extend this approach to 3-dimensional images in order to further refine this technology for possible clinical applications.

KW - chronic disease

KW - convolutional neural network

KW - machine learning

KW - neural network

KW - sinusitis

UR - http://www.scopus.com/inward/record.url?scp=85052390432&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85052390432&partnerID=8YFLogxK

U2 - 10.1002/alr.22196

DO - 10.1002/alr.22196

M3 - Article

C2 - 30098123

AN - SCOPUS:85052390432

JO - International Forum of Allergy and Rhinology

JF - International Forum of Allergy and Rhinology

SN - 2042-6976

ER -