[논문]딥러닝 기반 후두부 질환 내시경 영상판독 보조기술 개발

정인호; 황영준; 성의숙; 남경원

doi:10.9718/jber.2022.43.2.102

딥러닝 기반 후두부 질환 내시경 영상판독 보조기술 개발
Development of Deep Learning-based Clinical Decision Supporting Technique for Laryngeal Disease using Endoscopic Images 원문보기

Journal of biomedical engineering research : the official journal of the Korean Society of Medical & Biological Engineering, v.43 no.2, 2022년, pp.102 - 108

정인호 (부산대학교 의과대학 의공학협동과정) , 황영준 (부산대학교 의과대학 의공학협동과정) , 성의숙 (부산대학교 의과대학 이비인후과학교실) , 남경원 (부산대학교 의과대학 의공학협동과정)

Abstract ▼ AI-Helper

Purpose: To propose a deep learning-based clinical decision support technique for laryngeal disease on epiglottis, tongue and vocal cords. Materials and Methods: A total of 873 laryngeal endoscopic images were acquired from the PACS database of Pusan N ational University Yangsan Hospital. and VGG16 model was applied with transfer learning and fine-tuning. Results: The values of precision, recall, accuracy and F1-score for test dataset were 0.94, 0.97, 0.95 and 0.95 for epiglottis images, 0.91, 1.00, 0.95 and 0.95 for tongue images, and 0.90, 0.64, 0.73 and 0.75 for vocal cord images, respectively. Conclusion: Experimental results demonstrated that the proposed model have a potential as a tool for decision-supporting of otolaryngologist during manual inspection of laryngeal endoscopic images.

주제어

표/그림 (7)

그림 그림 1. 사용된 VGG16 모델의 기본구조 및 두 가지 전이학습 전략 (No-Fine-Tuned 모델, Fine-Tuned 모델) 개요도 Fig. 1. Basic structure of the utilized VGG16 model and two strategies of model transfer learning (No-Fine-Tuned Model and Fine-Tuned Model)
그림 그림 2. 후두덮개 영상증강 사례 (A) Original Image (B) Horizontal_Flip (C) Vertical_Flip (D) 60° Rotation Fig. 2. Example of epiglottis image augmentation (A) Original Image (B) Horizontal_Flip (C) Vertical_Flip (D) 60° Rotation
표 표 1. Epoch 값이 10, 20, 30, 40, 50인 경우 후두덮개, 혀, 성대영상에 대한 validation loss, validation accuracy 측정 결과(Fine-Tune Model; Validation dataset) Table 1. Variations in validation loss and validation accuracy at epoch 10, 20, 30, 40 and 50 for each target organ(Fine-Tuned Model; Validation dataset)
표 표 2. 입력데이터 및 모델 판정결과 예시(Fine-Tuned Model) Table 2. Examples of input image and the results of model output
표 표 3. 혼돈행렬 분석 결과(No-Fine-Tuned Model / Fine-Tuned Model) Table 3. Results of Confusion Matrix Analysis(No-Fine-Tuned Model / Fine-Tuned Model Order)
표 표 4. 정밀도, 재현율, 정확도, F1-값 계산 결과(No-Fine-Tuned Model / Fine-Tuned Model) Table 4. Values of precision, recall, accuracy and F1-score(No-Fine-Tuned Model / Fine-Tuned Model Order)
그림 그림 3. TDS "Transfer Learning from pre-trained models"에서 제시한 최적 가이드라인 모식도[14] Fig. 3. Scheme of optimal fine-tuning guideline from "Transfer Learning from pre-trained model" by TDS[14]

참고문헌 (15)

https://www.nhis.or.kr/nhis/together/wbhaea01600m01.do?modeview&articleNo13 8272 (Accessed on Dec 22, 2021)
Jung YS. Human papillomavirus in head and neck cancer: several questions. Korean J Otorhinolaryngol-Head Neck Surg 2014;57(3):143-150.
https://www.amc.seoul.kr/asan/healthinfo/disease/disease-Detail.do?contentId33914 (Accessed on Dec 22, 2021)
Sung J, Park S, Lee SM, Bae W, Park B, Jung E, et al. Added value of deep learning-based detection system for multiple major findings on chest radiographs: a randomized crossover study. Radiology 2021;299:450-459.

상세보기
Han SM, Hwang SI, Lee HJ. The classification of renal cancer in 3-phase CT images using a deep learning method. J Digit Imaging 2019;32(4):638-643.

상세보기
Suh CH, Shim WH, Kim SJ, Roh JH, Lee JH, Kim MJ, et al. Development and validation of a deep learning-based automatic brain segmentation and classification algorithm for Alzheimer disease using 3D T1-weighted volumetric images. Am J Neuroradiol 2020;41(12):2227-2234.

상세보기
Yoon HJ, Kim S, Kim JH, Keum JS, Oh SI, Jo J, et al. A lesion-based convolutional neural network improves endoscopic detection and depth prediction of early gastric cancer. J Clin Med 2019;8(9):1310.

상세보기
Jin EH. Lee D, Bae JH, Kang HY, Kwak MS, Seo JY, et al. Improved accuracy in optical diagnosis of colorectal polyps using convolutional neural networks with visual explanations. Gastroenterology, 2020;158(8):2169-2179.e8.
https://scienceon.kisti.re.kr/srch/selectPORSrchReport.do?cnTRKO202000004396 (Acc essed on Dec 22, 2021)
Amina B, Nadjia B, Azeddine B. Gastrointestinal image classification based on VGG16 and transfer learning. 2021 International Conference on Information Systems and Advanced Technologies (ICISAT). pp. 1-5.
Kim GH, Sung ES, Nam KW. Automated laryngeal mass detection algorithm for home-based self-screening test based on convolutional neural network. Biomed Eng Online 2021;20(51):1-10.
Yoo TK, Choi JY, Jang YI, Oh I, Ryu IH. Toward automated severe pharyngitis detection with smartphone camera using deep learning networks. Comput Biol Med 2020;125(103980):1-9.
Cho WK, Choi SH. Comparison of convolutional neural network models for determination of vocal fold normality in laryngoscopic images. J voice 2020;S0892-1997(20):30292-30297.
Cho WK, Lee YJ, Joo HA, Jeong IS, Choi Y, Nam SY, et al. Diagnostic accuracies of laryngeal disease using a convolutional neural network-based image classification system. Laryngoscope 2021;131(11):2558-2566.
https://towardsdatascience.com/transfer-learning-from-pretrained-models-f2393f124751 (Accessed on Dec 22, 2021)

내보내기 구분	파일저장 인쇄 메일전송
구성항목	기본정보 상세정보 관리번호, 논문명, 저널/프로시딩명, 저자 , 발행년, 권, 호, 시작페이지, 끝페이지, 발행기관 관리번호, 논문명, 대등논문명, 저자 , 저널/프로시딩명, 발행기관, 발행년, 발행언어, 권, 호, 시작페이지, 끝페이지, ISBN, ISSN, 주제분야, 키워드, 초록(한글), 초록(영문), 저자(소속기관)
저장형식	Text(ASCII format) Excel format RefWorks Direct Export RIS format (for Reference Manager, ProCite, EndNote), Scholar's Aids, Mendeley
메일정보	받는사람 (필수) @ 보내는사람 (선택) @ 제목 내용 KISTI 검색결과 이메일 서비스
안내	총 건의 자료가 검색되었습니다. 다운받으실 자료의 인덱스를 입력하세요. (1-10,000) 검색결과의 순서대로 최대 10,000건 까지 다운로드가 가능합니다. 데이타가 많을 경우 속도가 느려질 수 있습니다.(최대 2~3분 소요) 다운로드 파일은 UTF-8 형태로 저장됩니다. 파일의 내용이 제대로 보이지 않을실 때는 웹브라우저 상단의 보기 -> 인코딩 -> 자동선택 여부를 확인하십시오. ~ Text(ASCII format) Excel format

연합인증