Intelligent Diagnosis of Thyroid Ultrasound Imaging Using an Ensemble of Deep Learning Methods

被引:17
|
作者
Vasile, Corina Maria [1 ,2 ]
Udristoiu, Anca Loredana [3 ]
Ghenea, Alice Elena [4 ]
Popescu, Mihaela [5 ]
Gheonea, Cristian [6 ]
Niculescu, Carmen Elena [6 ]
Ungureanu, Anca Marilena [4 ]
Udristoiu, Stefan [3 ]
Drocas, Andrei Ioan [7 ]
Gruionu, Lucian Gheorghe [8 ]
Gruionu, Gabriel [9 ]
Iacob, Andreea Valentina [3 ]
Alexandru, Dragos Ovidiu [10 ]
机构
[1] Univ Med & Pharm Craiova, PhD Sch Dept, Craiova 200349, Romania
[2] Cty Clin Emergency Hosp Craiova, Dept Pediat Cardiol, Craiova 200642, Romania
[3] Univ Craiova, Fac Automat Comp & Elect, Craiova 200776, Romania
[4] Univ Med & Pharm Craiova, Dept Bacteriol Virol Parasitol, Craiova 200349, Romania
[5] Univ Med & Pharm Craiova, Dept Endocrinol, Craiova 200349, Romania
[6] Univ Med & Pharm Craiova, Dept Pediat, Craiova 200349, Romania
[7] Univ Med & Pharm Craiova, Dept Urol, Craiova 200349, Romania
[8] Univ Craiova, Fac Mech, Craiova 200512, Romania
[9] Indiana Univ Sch Med, Dept Med, Indianapolis, IN 46202 USA
[10] Univ Med & Pharm Craiova, Dept Med Informat & Biostat, Craiova 200349, Romania
来源
MEDICINA-LITHUANIA | 2021年 / 57卷 / 04期
关键词
thyroid disorders; ultrasound image; deep learning; neural networks; HASHIMOTOS-THYROIDITIS; NODULE; ULTRASONOGRAPHY; IMAGES;
D O I
10.3390/medicina57040395
中图分类号
R5 [内科学];
学科分类号
1002 ; 100201 ;
摘要
Background and Objectives: At present, thyroid disorders have a great incidence in the worldwide population, so the development of alternative methods for improving the diagnosis process is necessary. Materials and Methods: For this purpose, we developed an ensemble method that fused two deep learning models, one based on convolutional neural network and the other based on transfer learning. For the first model, called 5-CNN, we developed an efficient end-to-end trained model with five convolutional layers, while for the second model, the pre-trained VGG-19 architecture was repurposed, optimized and trained. We trained and validated our models using a dataset of ultrasound images consisting of four types of thyroidal images: autoimmune, nodular, micro-nodular, and normal. Results: Excellent results were obtained by the ensemble CNN-VGG method, which outperformed the 5-CNN and VGG-19 models: 97.35% for the overall test accuracy with an overall specificity of 98.43%, sensitivity of 95.75%, positive and negative predictive value of 95.41%, and 98.05%. The micro average areas under each receiver operating characteristic curves was 0.96. The results were also validated by two physicians: an endocrinologist and a pediatrician. Conclusions: We proposed a new deep learning study for classifying ultrasound thyroidal images to assist physicians in the diagnosis process.
引用
收藏
页数:14
相关论文
共 50 条
  • [1] Intelligent Ultrasound Imaging for Enhanced Breast Cancer Diagnosis: Ensemble Transfer Learning Strategies
    Rao, Kuncham Sreenivasa
    Terlapu, Panduranga Vital
    Jayaram, D.
    Raju, Kalidindi Kishore
    Kumar, G. Kiran
    Pemula, Rambabu
    Gopalachari, M. Venu
    Rakesh, S.
    IEEE ACCESS, 2024, 12 : 22243 - 22263
  • [2] Deep learning for intelligent diagnosis in thyroid scintigraphy
    Qiao, Tingting
    Liu, Simin
    Cui, Zhijun
    Yu, Xiaqing
    Cai, Haidong
    Zhang, Huijuan
    Sun, Ming
    Lv, Zhongwei
    Li, Dan
    JOURNAL OF INTERNATIONAL MEDICAL RESEARCH, 2021, 49 (01)
  • [3] Thyroid nodule classification in ultrasound imaging using deep transfer learning
    Xu, Yan
    Xu, Mingmin
    Geng, Zhe
    Liu, Jie
    Meng, Bin
    BMC CANCER, 2025, 25 (01)
  • [4] A novel intelligent thyroid nodule diagnosis system over ultrasound images based on deep learning
    Yi, Zhike
    Hao, Aimin
    Song, Wenfeng
    Li, Hongyi
    Li, Bowen
    2017 INTERNATIONAL CONFERENCE ON VIRTUAL REALITY AND VISUALIZATION (ICVRV 2017), 2017, : 150 - 155
  • [5] Intelligent Imaging Technology in Diagnosis of Colorectal Cancer Using Deep Learning
    Yang, Tao
    Liang, Ning
    Li, Jing
    Yang, Yi
    Li, Yuedan
    Huang, Qian
    Li, Renzhi
    He, Xianli
    Zhang, Hongxin
    IEEE ACCESS, 2019, 7 : 178839 - 178847
  • [6] Ultrasound image analysis using deep learning algorithm for the diagnosis of thyroid nodules
    Song, Junho
    Chai, Young Jun
    Masuoka, Hiroo
    Park, Sun-Won
    Kim, Su-jin
    Choi, June Young
    Kong, Hyoun-Joong
    Lee, Kyu Eun
    Lee, Joongseek
    Kwak, Nojun
    Yi, Ka Hee
    Miyauchi, Akira
    MEDICINE, 2019, 98 (15)
  • [7] DIAGNOSIS OF CUBITAL TUNNEL SYNDROME BY ULTRASOUND IMAGING USING DEEP LEARNING
    Inui, A.
    Nishimoto, H.
    Mifune, Y.
    Mukouhara, S.
    Yoshikawa, T.
    Shinohara, I
    Kato, T.
    Furukawa, T.
    Kuroda, R.
    JOURNAL OF THE PERIPHERAL NERVOUS SYSTEM, 2022, 27 : S17 - S18
  • [8] Diagnosis of Developmental Dysplasia of the Hip by Ultrasound Imaging Using Deep Learning
    Kinugasa, Maki
    Inui, Atsuyuki
    Satsuma, Shinichi
    Kobayashi, Daisuke
    Sakata, Ryosuke
    Morishita, Masayuki
    Komoto, Izumi
    Kuroda, Ryosuke
    JOURNAL OF PEDIATRIC ORTHOPAEDICS, 2023, 43 (07) : E538 - E544
  • [9] Differential Diagnosis of Benign and Malignant Thyroid Nodules Using Deep Learning Radiomics of Thyroid Ultrasound Images
    Zhou, Hui
    Jin, Yinhua
    Dai, Lei
    Zhang, Meiwu
    Qiu, Yuqin
    Wang, Kun
    Tian, Jie
    Zheng, Jianjun
    EUROPEAN JOURNAL OF RADIOLOGY, 2020, 127
  • [10] Ensemble Deep Learning Model for Multicenter Classification of Thyroid Nodules on Ultrasound Images
    Wei, Xi
    Gao, Ming
    Yu, Ruiguo
    Liu, Zhiqiang
    Gu, Qing
    Liu, Xun
    Zheng, Zhiming
    Zheng, Xiangqian
    Zhu, Jialin
    Zhang, Sheng
    MEDICAL SCIENCE MONITOR, 2020, 26