EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Detecting Macular Disease Based on Optical Coherence Tomography Using a Deep Convolutional Network

Jinyoung Han iD, Seong Choi, Ji In Park iD, Joon Seo Hwang, Jeong Mo Han iD, Junseo Ko, Jeewoo Yoon, Daniel Duck-Jin Hwang iD

Journal of Clinical Medicine · 28.01.2023 · 12 (3) · 1005

DOI: 10.3390/jcm12031005 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Clinical Medicine, online erschienen am 28.01.2023. Verfasst von Jinyoung Han und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Applied Artificial Intelligence, Sungkyunkwan University, Seoul 03063, Republic of Korea. Die Arbeit wurde bisher 23-mal zitiert (Crossref) und nennt 20 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Research Foundation of Korea.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Neovascular age-related macular degeneration (nAMD) and central serous chorioretinopathy (CSC) are two of the most common macular diseases. This study proposes a convolutional neural network (CNN)-based deep learning model for classifying the subtypes of nAMD (polypoidal choroidal vasculopathy, retinal angiomatous proliferation, and typical nAMD) and CSC (chronic CSC and acute CSC) and healthy individuals using single spectral–domain optical coherence tomography (SD–OCT) images. The proposed model was trained and tested using 6063 SD–OCT images from 521 patients and 47 healthy participants. We used three well-known CNN architectures (VGG–16, VGG–19, and ResNet) and two customized classification layers. Additionally, transfer learning and mix–up-based data augmentation were applied to improve robustness and accuracy. Our model demonstrated high accuracies of 99.7% and 91.1% in the nAMD and CSC classification and retinopathy (nAMD and CSC) subtype classification, including normal participants, respectively. Furthermore, we performed an external test to compare the classification accuracy with that of eight ophthalmologists, and our model showed the highest accuracy. The region determined to be important for classification by the model was confirmed using gradient-weighted class activation mapping. The model’s clinical criteria were similar to that of the ophthalmologists.

Schlagwörter

Stichwörter der Autor:innen

convolutional neural networkdeep learningmedical imageretinopathy

Autor:innen und Institutionen

  • Department of Applied Artificial Intelligence, Sungkyunkwan University, Seoul 03063, Republic of Korea.
  • Department of Human-Artificial Intelligence Interaction, Sungkyunkwan University, Seoul 03063, Republic of Korea.
  • RaonData, Seoul 04615, Republic of Korea.
  • Department of Medicine, Kangwon National University Hospital, Kangwon National University School of Medicine, Chuncheon 24341, Gangwon-do, Republic of Korea.
  • Seoul Plus Eye Clinic, Seoul 01751, Republic of Korea.
  • Seoul Bombit Eye Clinic, Sejong 30127, Republic of Korea.

ORCID-Profile: Jinyoung Han, Ji In Park, Jeong Mo Han, Daniel Duck-Jin Hwang

Förderung

  • National Research Foundation of Korea

Zitieren

APA

Han, J., Choi, S., Park, J. I., Hwang, J. S., Han, J. M., Ko, J., Yoon, J., & Hwang, D. D. (2023). Detecting Macular Disease Based on Optical Coherence Tomography Using a Deep Convolutional Network. Journal of Clinical Medicine, 12(3), 1005. https://doi.org/10.3390/jcm12031005

Vancouver

Han J, Choi S, Park JI, Hwang JS, Han JM, Ko J, et al. Detecting Macular Disease Based on Optical Coherence Tomography Using a Deep Convolutional Network. Journal of Clinical Medicine. 2023;12(3):1005. doi:10.3390/jcm12031005

BibTeX
@article{Han2023,
  author = {Han, Jinyoung and Choi, Seong and Park, Ji In and Hwang, Joon Seo and Han, Jeong Mo and Ko, Junseo and Yoon, Jeewoo and Hwang, Daniel Duck-Jin},
  title = {Detecting Macular Disease Based on Optical Coherence Tomography Using a Deep Convolutional Network},
  journal = {Journal of Clinical Medicine},
  year = {2023},
  volume = {12},
  number = {3},
  pages = {1005},
  doi = {10.3390/jcm12031005}
}
Kostenlos starten →