딥러닝 모델을 이용한 휴대용 무선 초음파 영상에서의 경동맥 내중막 두께 자동 분할 알고리즘 개발
Development of Automatic Segmentation Algorithm of Intima-media Thickness of Carotid Artery in Portable Ultrasound Image Based on Deep Learning
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Measuring Intima-media thickness (IMT) with ultrasound images can help early detection of coronary artery disease. As a result, numerous machine learning studies have been conducted to measure IMT. However, most of these studies require several steps of pre-treatment to extract the boundary, and some require manual inter�vention, so they are not suitable for on-site treatment in urgent situations. in this paper, we propose to use deep learning networks U-Net, Attention U-Net, and Pretrained U-Net to automatically segment the intima-media com�plex. This study also applied the HE, HS, and CLAHE preprocessing technique to wireless portable ultrasound diag�nostic device images. As a result, The average dice coefficient of HE applied Models is 71% and CLAHE applied Models is 70%, while the HS applied Models have improved as 72% dice coefficient. Among them, Pretrained U�Net showed the highest performance with an average of 74%. When comparing this with the mean value of IMT measured by Conventional wired ultrasound equipment, the highest correlation coefficient value was shown in the HS applied pretrained U-Net.
Abstract
Measuring Intima-media thickness (IMT) with ultrasound images can help early detection of coronary artery disease. As a result, numerous machine learning studies have been conducted to measure IMT. However, most of these studies require several steps of pre-treatment to extract the boundary, and some require manual inter�vention, so they are not suitable for on-site treatment in urgent situations. in this paper, we propose to use deep learning networks U-Net, Attention U-Net, and Pretrained U-Net to automatically segment the intima-media com�plex. This study also applied the HE, HS, and CLAHE preprocessing technique to wireless portable ultrasound diag�nostic device images. As a result, The average dice coefficient of HE applied Models is 71% and CLAHE applied Models is 70%, while the HS applied Models have improved as 72% dice coefficient. Among them, Pretrained U�Net showed the highest performance with an average of 74%. When comparing this with the mean value of IMT measured by Conventional wired ultrasound equipment, the highest correlation coefficient value was shown in the HS applied pretrained U-Net.
참고문헌 (23)
- Honda O, "Echolucent carotid plaques predict future coronary events in patients with coronary artery disease", Journal of the American College of Cardiology, 43(7), pp.1177~1184 (2004)
- Kwon T, "Measurement of Carotid Intima-Media Thickness in ultrasound images by means of an automatic segmentation process based on machine learning", The Korean Society of Circulation, 37, pp.103~107 (2007)
- Liang Q, "A multiscale dynamic programming procedure for boundary detection in ultrasonic artery images", IEEE Transactions Medical Imaging, 19, pp.127~142 (2000)
- Kim JS, "The significance of muscle echo intensity on ultrasound for focal neuropathy: The median- to ulnar-innervated muscle echo intensity ratio in carpal tunnel syndrome", Journal of Clinical Neurophysiology, 127, pp.880~885 (2016)
- Litjens G, "A survey on deep learning in medical image analysis", Medical Image Analysis, 42, pp.60~88 (2017)
- Ravì D, "Deep Learning for Health Informatics", IEEE Journal of Biomedical and Health Informatics, 21(1), pp.4~21 (2017)
- Li Q, "An improved approach for accurate and efficient measurement of common carotid artery intima-media thickness in ultrasound images", Journal of Biomedical Research International, 2014 (2014)
- Loizou CP, "Manual and automated media and intima thickness measurements of the common carotid artery", IEEE Transactions Medical Imaging, 56(5) (2009)
- Li H, "Ultrasound intima-media thickness measurement of the carotid artery using ant colony optimization combined with a curvelet-based orientation-selective filter", Medical Physics, 43(4), pp.1795~1807 (2016)
- Sudha S, "Convolutional Neural Network for Segmentation and Measurement of Intima Media Thickness", Journal of Medical Systems, 42(154) (2018)
- Liang Q, "A multiscale dynamic programming procedure for boundary detection in ultrasonic artery images", IEEE Transactions Medical Imaging, 19(2), pp.127~142 (2000)
- 이상민, "폐암검진에서 인공지능 기술의 활용", 대한영상의학회지, 80(5), pp.872~879 (2019), 대한영상의학회[DOI]
- Hyun DW, "Measurement of the Carotid Intima, Media and Intima-Media Thickness with Ultrasound and New Software", The Korean Society of Circulation, 35(8), pp.625~632 (2005)
- Coppola G, "The impact of the ketogenic diet on arterial morphology and endothelial function in children and young adults with epilepsy : A case–control study", Seizure, 23(4), pp.260~265 (2014)
- Kim DJ, "Image contrast enhancement using K-means algorithm", Master, pp.2008
- 강현우, "히스토그램 보정을 통한 적응형 명암비 향상 방법", 한국산학기술학회논문지, 11(3), pp.958~964 (2010), 한국산학기술학회
- Ronneberger O, "U-Net: Convolutional Networks for Biomedical Image Segmentation", Medical Image Computing and Computer-Assisted Intervention, 9351, pp.234~241 (2015)
- Schlemper J, "Attention gated networks: Learning to leverage salient regions in medical images", Medical Image Analysis, 53, pp.197~207 (2019)
- Oktay O, "Attention U-Net: Learning Where to Look for the Pancreas", ArXiv, ~ (2018)
- Weiss K, "A Survey on Transfer Learning", Journal of Big Data, 3(9), pp.1~40 (2014)
- He K, "Deep residual learning for image recognition", Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp.770~778 (2016)
- Yang HS, "Carotid Ultrasound: Imaging and Interpretation for Clinicians", Clinical Ultrasound, 3, pp.1~7 (2018)
- Lee SJ, "Extracranial Carotid Duplex Ultrasonography. Part I - Basic Principles and Standard Examination for Carotid and Vertebral Arteries, and Jugular Veins", Journal of the Neurosonol Neuroimag, 10(2), pp.47~60 (2018)