Universal Lesion Detection and Classification Using Limited Data and Weakly-Supervised Self-training
文献类型: 会议论文
第一作者: Varun Naga
作者: Varun Naga 1 ; Tejas Sudharshan Mathai 1 ; Angshuman Paul 2 ; Ronald M. Summers 1 ;
作者机构: 1.Imaging Biomarkers and Computer-Aided Diagnosis Laboratory, Radiology and Imaging Sciences, Clinical Center, National Institutes of Health
2.Indian Institute of Technology
关键词: CT;Detection;Classification;Deep learning;Self-training
会议名称: International Workshop on Medical Image Learning with Limited and Noisy Data
主办单位:
页码: 55-64
摘要: Radiologists identify, measure, and classify clinically significant lesions routinely for cancer staging and tumor burden assessment. As these tasks are repetitive and cumbersome, only the largest lesion is identified leaving others of potential importance unmentioned. Automated deep learning-based methods for lesion detection have been proposed in literature to help relieve their tasks with the publicly available DeepLesion dataset (32,735 lesions, 32,120 CT slices, 10,594 studies, 4,427 patients, 8 body part labels). However, this dataset contains missing lesions, and displays a severe class imbalance in the labels. In our work, we use a subset of the DeepLesion dataset (boxes + tags) to train a state-of-the-art VFNet model to detect and classify suspicious lesions in CT volumes. Next, we predict on a larger data subset (containing only bounding boxes) and identify new lesion candidates for a weakly-supervised self-training scheme. The self-training is done across multiple rounds to improve the model's robustness against noise. Two experiments were conducted with static and variable thresholds during self-training, and we show that sensitivity improves from 72.5% without self-training to 76.4% with self-training. We also provide a structured reporting guideline through a "Lesions" subsection for entry into the "Findings" section of a radiology report. To our knowledge, we are the first to propose a weakly-supervised self-training approach for joint lesion detection and tagging in order to mine for under-represented lesion classes in the DeepLesion dataset.
分类号: tp391.41-53
- 相关文献
[1]Detection of Lymph Nodes in T2 MRI Using Neural Network Ensembles. Zhiyong Lu,Ronald M. Summers,Tejas Sudharshan Mathai,Sungwon Lee,Daniel C. Elton,Thomas C. Shen,Yifan Peng. 2021
[2]Class Imbalance Correction for Improved Universal Lesion Detection and Tagging in CT. Peter D. Erickson,Tejas Sudharshan Mathai,Ronald M. Summers. 2022
[3]实时荧光定量PCR在植物害虫研究中的应用. 王忠跃,孔繁芳,武艳霞. 2011
[4]CT扫描分析秸秆隔层孔隙特征及其对土壤水入渗的影响. 张宏媛,逄焕成,卢闯,刘娜,张晓丽,李玉义. 2019
[5]适合马铃薯块茎计算机断层扫描原位观察的栽培基质筛选. 谭杰,刘建刚,刘菊,简银巧,李广存,金黎平,徐建飞. 2023
[6]Learning Functional Embedding of Genes Governed by Pair-wised Labels. Cao, Jingjun,Wu, Zhengli,Ye, Wenting,Wang, Haohan. 2017
[7]A Walking Assistive Device with Intention Detection using Back-driven Pneumatic Artificial Muscles. Takahiro Kanno,Daisuke Morisaki,Ryoken Miyazaki,Gen Endo,Kenji Kawashima. 2015
[8]Classification Methods of Remote Sensing Image Based on Decision Tree Technologies. Jiang, Lihua,Wang, Wensheng,Yang, Xiaorong,Xie, Nengfu,Jiang, Lihua,Wang, Wensheng,Yang, Xiaorong,Xie, Nengfu,Cheng, Youping. 2011
[9]Classification of the urbanization types in the Yangtze River Delta. Xu, Juan,Yang, Li,Zhang, Wei. 2012
[10]Non-specific lipid transfer proteins in plants: presenting new advances and an integrated functional analysis. Liu, Fang,Lu, Changming,Zeng, Xinhua,Li, Yunjing,Wu, Gang,Zhang, Xiaobo,Fu, Donghui.
[11]The Feasibility and Stability of Distinguishing the Kiwi Fruit Geographical Origin Based on Electronic Nose Analysis. Ma, Yiyan,Guo, Boli,Wei, Yimin,Wei, Shuai,Zhao, Haiyan.
[12]Flavonoid glycosides as floral origin markers to discriminate of unifloral bee pollen by LC-MS/MS. Zhou, Jinhui,Wu, Liming,Li, Yi,Sun, Liping,Zhou, Jinhui,Wu, Liming,Li, Yi,Zhou, Jinhui,Wu, Liming,Li, Yi,Qi, Yitao,Ritho, Joan,Zhang, Yongxin,Zheng, Xiaowei,Sun, Liping.
[13]Genome-Wide Identification and Classification of Soybean C2H2 Zinc Finger Proteins and Their Expression Analysis in Legume-Rhizobium Symbiosis. Yuan, Songli,Li, Rong,Zhang, Chanjuan,Chen, Limiao,Hao, Qingnan,Zhang, Xiaojuan,Chen, Haifeng,Shan, Zhihui,Yang, Zhonglu,Chen, Shuilian,Qiu, Dezhen,Zhou, Xinan,Yuan, Songli,Li, Rong,Zhang, Chanjuan,Chen, Limiao,Hao, Qingnan,Zhang, Xiaojuan,Chen, Haifeng,Shan, Zhihui,Yang, Zhonglu,Chen, Shuilian,Qiu, Dezhen,Zhou, Xinan,Li, Xiangyong,Wang, Lei,Ke, Danxia. 2018
[14]Hyper-spectral characteristics and classification of farmland soil in northeast of China. Lu Yan-li,Bai You-lu,Yang Li-ping,Wang Lei,Wang Yi-lun,Ni Lu,Zhou Li-ping. 2015
[15]Similarity-Based Classification for Big Non-Structured and Semi-Structured Recipe Data. Chen, Wei,Zhao, Xiangyu,Chen, Wei,Zhao, Xiangyu. 2016
[16]Feature Extraction and Classification of Animal Blood Spectra with Support Vector Machine. Lu Peng-fei,Fan Ya,Zhou Lin-hua,Gao Bin,Qian Jun,Liu Lin-na,Zhao Si-yan,Kong Zhi-feng. 2017
[17]An Enhanced IT2FCM*Algorithm Integrating Spectral Indices and Spatial Information for Multi-Spectral Remote Sensing Image Clustering. Guo, Jifa,Huo, Hongyuan. 2017
[18]Genome-wide identification and evolutionary and expression analyses of MYB-related genes in land plants. Du, Hai,Wang, Yong-Bin,Xie, Yi,Zhang, Shuang-Shuang,Huang, Yu-Bi,Du, Hai,Tang, Yi-Xiong,Liang, Zhe,Jiang, San-Jie.
[19]Characterization of volatile components in four vegetable oils by headspace two-dimensional comprehensive chromatography time-of-flight mass spectrometry. Hu, Wei,Zhang, Liangxiao,Li, Peiwu,Wang, Xiupin,Zhang, Qi,Xu, Baocheng,Sun, Xiaoman,Ma, Fei,Ding, Xiaoxia,Hu, Wei,Zhang, Qi,Li, Peiwu,Zhang, Qi,Ma, Fei,Zhang, Liangxiao,Li, Peiwu,Ding, Xiaoxia,Zhang, Liangxiao,Li, Peiwu,Wang, Xiupin,Xu, Baocheng,Sun, Xiaoman,Ma, Fei.
[20]Detecting peanuts inoculated with toxigenic and atoxienic Aspergillus flavus strains with fluorescence hyperspectral imagery. Xing, Fuguo,Yao, Haibo,Hruska, Zuzana,Kincaid, Russell,Zhu, Fengle,Xing, Fuguo,Liu, Yang,Xing, Fuguo,Brown, Robert L.,Bhatnagar, Deepak. 2017
作者其他论文 更多>>
-
Automated Measurement of Pericoronary Adipose Tissue Attenuation and Volume in Ct Angiography
作者:Andrew M. Nguyen;Tejas Sudharshan Mathai;Liangchen Liu;Jianfei Liu;Ronald M. Summers
关键词:Three-dimensional displays;Volume measurement;Angiography;Biomedical measurement;Attenuation measurement;Attenuation;Fats
-
Automated Classification of Multi-Parametric Body MRI Series
作者:Boah Kim;Tejas Sudharshan Mathai;Kimberly Helm;Ronald M. Summers
关键词:Protocols;Sensitivity;Automation;Magnetic resonance imaging;Radiology;Pelvis;DICOM
-
Semantic Image Synthesis for Abdominal CT
作者:Yan Zhuang;Benjamin Hou;Tejas Sudharshan Mathai;Pritam Mukherjee;Boah Kim;Ronald M. Summers
关键词:CT;Abdomen;Diffusion model;Semantic image synthesis
-
Anatomical Location-Guided Deep Learning-Based Genetic Cluster Identification of Pheochromocytomas and Paragangliomas from CT Images
作者:Bikash Santra;Abhishek Jha;Pritam Mukherjee;Mayank Patel;Karel Pacak;Ronald M. Summers
关键词:Pheochromocytoma;Paraganglioma;Genetic mutations;Transformer;CT images;Deep learning;Radiogenomics;PPGL
-
Improving Automatic Segmentation of lymphoma with Additional Medical Knowledge Priors
作者:Liangchen Liu;Benjamin Hou;Manas Kumar Nag;Jianfei Liu;Navid Hasani;Thomas C. Shen;Yingying Zhu;Babak Saboury;Jing Xiao;Ronald M. Summers
关键词:Training;Image segmentation;Shape;Computed tomography;Pipelines;Positron emission tomography;Radiotracer
-
A Dual-Branch Network with Mixed and Self-Supervision for Medical Image Segmentation: An Application to Segment Edematous Adipose Tissue
作者:Jianfei Liu;Omid Shafaat;Ronald M. Summers
关键词:Self-supervision;Mixed supervision;Dual-branch;Residual u-Net;Edema
-
A Dual-Branch Network with Mixed and Self-Supervision for Medical Image Segmentation: An Application to Segment Edematous Adipose Tissue
作者:Jianfei Liu;Omid Shafaat;Ronald M. Summers
关键词:Self-supervision;Mixed supervision;Dual-branch;Residual U-Net;Edema


