EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Deep learning algorithm for the automated detection and classification of nasal cavity mass in nasal endoscopic images

Kyung Won Kwon, Seong Hyeon Park, Dong Hoon Lee, Dong-Young Kim, Il-Ho Park, Hyun-Jin Cho, Jong Seung Kim, Joo Yeon Kim, Sang Duk Hong, Shin Ae Kim, Shin Hyuk Yoo iD, Soo Kyoung Park und 9 weitere

PLOS ONE · 13.03.2024 · 19 (3) · e0297536

DOI: 10.1371/journal.pone.0297536 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS ONE, online erschienen am 13.03.2024. Verfasst von Kyung Won Kwon und 20 weiteren Autor:innen. Beteiligte Einrichtung: Department of Otolaryngology, Samsung Changwon Hospital, Sungkyunkwan University School of Medicine, Changwon, Korea. Hauptthemen laut MEDLINE: Deep Learning, Neoplasms. Die Arbeit wurde bisher 17-mal zitiert (Crossref) und nennt 45 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Korean Rhinologic Society.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Nasal endoscopy is routinely performed to distinguish the pathological types of masses. There is a lack of studies on deep learning algorithms for discriminating a wide range of endoscopic nasal cavity mass lesions. Therefore, we aimed to develop an endoscopic-examination-based deep learning model to detect and classify nasal cavity mass lesions, including nasal polyps (NPs), benign tumors, and malignant tumors. The clinical feasibility of the model was evaluated by comparing the results to those of manual assessment. Biopsy-confirmed nasal endoscopic images were obtained from 17 hospitals in South Korea. Here, 400 images were used for the test set. The training and validation datasets consisted of 149,043 normal nasal cavity, 311,043 NP, 9,271 benign tumor, and 5,323 malignant tumor lesion images. The proposed Xception architecture achieved an overall accuracy of 0.792 with the following class accuracies on the test set: normal = 0.978 ± 0.016, NP = 0.790 ± 0.016, benign = 0.708 ± 0.100, and malignant = 0.698 ± 0.116. With an average area under the receiver operating characteristic curve (AUC) of 0.947, the AUC values and F1 score were highest in the order of normal, NP, malignant tumor, and benign tumor classes. The classification performances of the proposed model were comparable with those of manual assessment in the normal and NP classes. The proposed model outperformed manual assessment in the benign and malignant tumor classes (sensitivities of 0.708 ± 0.100 vs. 0.549 ± 0.172, 0.698 ± 0.116 vs. 0.518 ± 0.153, respectively). In urgent (malignant) versus nonurgent binary predictions, the deep learning model achieved superior diagnostic accuracy. The developed model based on endoscopic images achieved satisfactory performance in classifying four classes of nasal cavity mass lesions, namely normal, NP, benign tumor, and malignant tumor. The developed model can therefore be used to screen nasal cavity lesions accurately and rapidly.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningNeoplasmsHumansNasal CavityAlgorithmsEndoscopy

Autor:innen und Institutionen

  • Department of Otolaryngology, Samsung Changwon Hospital, Sungkyunkwan University School of Medicine, Changwon, Korea.
  • Department of Biomedical Informatics, College of Medicine, Konyang University, Daejeon, Korea.
  • Department of Otolaryngology-Head and Neck Surgery, Chonnam National University Medical School & Hwasun Hospital, Hwasun, Korea.
  • Department of Otorhinolaryngology-Head and Neck Surgery, Seoul National University Hospital, Seoul, Korea.
  • Department of Otorhinolaryngology-Head and Neck Surgery, Korea University Guro Hospital, Korea University College of Medicine, Seoul, Korea.
  • Department of Otorhinolaryngology, Gyeongsang National University School of Medicine and Gyeongsang National University Hospital, Jinju, Korea.

ORCID-Profile: Shin Hyuk Yoo, Jong-Yeup Kim, Myeong Sang Yu

Förderung

  • Korean Rhinologic Society

Zitieren

APA

Kwon, K. W., Park, S. H., Lee, D. H., Kim, D. Y., Park, I. H., Cho, H. J., Kim, J. S., Kim, J. Y., Hong, S. D., Kim, S. A., Yoo, S. H., Park, S. K., Heo, S. J., Kim, S. H., Won, T. B., Choi, W. R., Kim, Y. M., Kim, Y. W., Kim, J. Y., …, & Yu, M. S. (2024). Deep learning algorithm for the automated detection and classification of nasal cavity mass in nasal endoscopic images. PLOS ONE, 19(3), e0297536. https://doi.org/10.1371/journal.pone.0297536

Vancouver

Kwon KW, Park SH, Lee DH, Kim DY, Park IH, Cho HJ, et al. Deep learning algorithm for the automated detection and classification of nasal cavity mass in nasal endoscopic images. PLOS ONE. 2024;19(3):e0297536. doi:10.1371/journal.pone.0297536

BibTeX
@article{Kwon2024,
  author = {Kwon, Kyung Won and Park, Seong Hyeon and Lee, Dong Hoon and Kim, Dong-Young and Park, Il-Ho and Cho, Hyun-Jin and Kim, Jong Seung and Kim, Joo Yeon and Hong, Sang Duk and Kim, Shin Ae and Yoo, Shin Hyuk and Park, Soo Kyoung and Heo, Sung Jae and Kim, Sung Hee and Won, Tae-Bin and Choi, Woo Ri and Kim, Yong Min and Kim, Yong Wan and Kim, Jong-Yeup and Kwon, Jae Hwan and Yu, Myeong Sang},
  title = {Deep learning algorithm for the automated detection and classification of nasal cavity mass in nasal endoscopic images},
  journal = {PLOS ONE},
  year = {2024},
  volume = {19},
  number = {3},
  pages = {e0297536},
  doi = {10.1371/journal.pone.0297536}
}
Kostenlos starten →