EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Deep generative abnormal lesion emphasization validated by nine radiologists and 1000 chest X-rays with lung nodules

Shouhei Hanaoka iD, Yukihiro Nomura iD, Naoto Hayashi, Issei Sato, Soichiro Miki, Takeharu Yoshikawa, Hisaichi Shibata, Takahiro Nakao iD, Tomomi Takenaga, Hiroaki Koyama, Shinichi Cho, Noriko Kanemaru und 6 weitere

PLOS ONE · 2024

DOI: 10.1371/journal.pone.0315646 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS ONE, 2024. Verfasst von Shouhei Hanaoka und 17 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiology and Diagnostic Radiology and Preventive Medicine, The University of Tokyo Hospital, Bunkyo-ku, Tokyo, Japan. Hauptthemen laut MEDLINE: Radiography, Thoracic, Radiologists, Lung Neoplasms. Die Arbeit nennt 47 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Japan Society for the Promotion of Science KAKENHI, the Joint Usage/Research Center for Interdisciplinary Large-Scale Information Infrastructures and High-Performance Computing Infrastructure Projects in Japan, Siemens Healthcare KK.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

A general-purpose method of emphasizing abnormal lesions in chest radiographs, named EGGPALE (Extrapolative, Generative and General-Purpose Abnormal Lesion Emphasizer), is presented. The proposed EGGPALE method is composed of a flow-based generative model and L-infinity-distance-based extrapolation in a latent space. The flow-based model is trained using only normal chest radiographs, and an invertible mapping function from the image space to the latent space is determined. In the latent space, a given unseen image is extrapolated so that the image point moves away from the normal chest X-ray hyperplane. Finally, the moved point is mapped back to the image space and the corresponding emphasized image is created. The proposed method was evaluated by an image interpretation experiment with nine radiologists and 1,000 chest radiographs, of which positive suspected lung cancer cases and negative cases were validated by computed tomography examinations. The sensitivity of EGGPALE-processed images showed +0.0559 average improvement compared with that of the original images, with -0.0192 deterioration of average specificity. The area under the receiver operating characteristic curve of the ensemble of nine radiologists showed a statistically significant improvement. From these results, the feasibility of EGGPALE for enhancing abnormal lesions was validated. Our code is available at https://github.com/utrad-ical/Eggpale .

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Radiography, ThoracicRadiologistsLung NeoplasmsHumansTomography, X-Ray ComputedRadiographic Image Interpretation, Computer-AssistedROC CurveAlgorithmsDeep LearningSolitary Pulmonary Nodule

Autor:innen und Institutionen

  • Department of Radiology and Diagnostic Radiology and Preventive Medicine, The University of Tokyo Hospital, Bunkyo-ku, Tokyo, Japan.
  • Department of Computational Diagnostic Radiology and Preventive Medicine, The University of Tokyo Hospital, Bunkyo-ku, Tokyo, Japan.
  • Center for Frontier Medical Engineering, Chiba University, Chiba, Japan.
  • Department of Computer Science, Graduate School of Information Science and Technology, The University of Tokyo, Bunkyo-ku, Japan.
  • Toranomon Hospital, Minato-ku, Tokyo, Japan.
  • Kanto Rosai Hospital, Kawasaki City, Kanagawa Prefecture, Japan.

ORCID-Profile: Shouhei Hanaoka, Yukihiro Nomura, Takahiro Nakao, Hirotaka Matsuzaki, Osamu Abe

Förderung

  • Japan Society for the Promotion of Science KAKENHI
  • the Joint Usage/Research Center for Interdisciplinary Large-Scale Information Infrastructures and High-Performance Computing Infrastructure Projects in Japan
  • Siemens Healthcare KK
  • HIMEDIC Inc.

Zitieren

APA

Hanaoka, S., Nomura, Y., Hayashi, N., Sato, I., Miki, S., Yoshikawa, T., Shibata, H., Nakao, T., Takenaga, T., Koyama, H., Cho, S., Kanemaru, N., Fujimoto, K., Sakamoto, N., Nishiyama, T., Matsuzaki, H., Yamamichi, N., & Abe, O. (2024). Deep generative abnormal lesion emphasization validated by nine radiologists and 1000 chest X-rays with lung nodules. PLOS ONE. https://doi.org/10.1371/journal.pone.0315646

Vancouver

Hanaoka S, Nomura Y, Hayashi N, Sato I, Miki S, Yoshikawa T, et al. Deep generative abnormal lesion emphasization validated by nine radiologists and 1000 chest X-rays with lung nodules. PLOS ONE. 2024. doi:10.1371/journal.pone.0315646

BibTeX
@article{Hanaoka2024,
  author = {Hanaoka, Shouhei and Nomura, Yukihiro and Hayashi, Naoto and Sato, Issei and Miki, Soichiro and Yoshikawa, Takeharu and Shibata, Hisaichi and Nakao, Takahiro and Takenaga, Tomomi and Koyama, Hiroaki and Cho, Shinichi and Kanemaru, Noriko and Fujimoto, Kotaro and Sakamoto, Naoya and Nishiyama, Tomoya and Matsuzaki, Hirotaka and Yamamichi, Nobutake and Abe, Osamu},
  title = {Deep generative abnormal lesion emphasization validated by nine radiologists and 1000 chest X-rays with lung nodules},
  journal = {PLOS ONE},
  year = {2024},
  doi = {10.1371/journal.pone.0315646}
}
Kostenlos starten →