{"title":"基于深度卷积神经网络的建筑变形检测","authors":"S. Kulkarni, Rinku Rabidas","doi":"10.1109/SILCON55242.2022.10028896","DOIUrl":null,"url":null,"abstract":"Breast cancer is one of the threatening diseases among women throughout the world. The early detection is the only way to cure from cancer. Architectural distortion (AD) is one of the earliest symptoms of breast cancer which is mostly malignant in nature. Computer-aided detection (CAD) and particularly deep learning (DL) gives prominent solution for the detection and diagnosis of breast cancer. This paper presents a deep convolutional neural network (DCNN) architecture designed for the automatic detection of AD in digital mammography images. The proposed deep learning based model consists of series combination of down sampler and ResNet blocks. Due to stacking of these blocks, these layers learn more complex features which help to improved in sensitivity and performance of the model. A total of 150 mammograms are considered for experimentation purpose from publicly available dataset namely, DDSM. Hence the best result obtained in the proposed approach with Leave-One-Out cross validation technique, in terms of true positive rate 86% at 0.42 false positives per image (FPs/I).","PeriodicalId":183947,"journal":{"name":"2022 IEEE Silchar Subsection Conference (SILCON)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Detection of Architectural Distortion using Deep Convolutional Neural Network\",\"authors\":\"S. Kulkarni, Rinku Rabidas\",\"doi\":\"10.1109/SILCON55242.2022.10028896\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Breast cancer is one of the threatening diseases among women throughout the world. The early detection is the only way to cure from cancer. Architectural distortion (AD) is one of the earliest symptoms of breast cancer which is mostly malignant in nature. Computer-aided detection (CAD) and particularly deep learning (DL) gives prominent solution for the detection and diagnosis of breast cancer. This paper presents a deep convolutional neural network (DCNN) architecture designed for the automatic detection of AD in digital mammography images. The proposed deep learning based model consists of series combination of down sampler and ResNet blocks. Due to stacking of these blocks, these layers learn more complex features which help to improved in sensitivity and performance of the model. A total of 150 mammograms are considered for experimentation purpose from publicly available dataset namely, DDSM. Hence the best result obtained in the proposed approach with Leave-One-Out cross validation technique, in terms of true positive rate 86% at 0.42 false positives per image (FPs/I).\",\"PeriodicalId\":183947,\"journal\":{\"name\":\"2022 IEEE Silchar Subsection Conference (SILCON)\",\"volume\":\"10 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2022-11-04\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2022 IEEE Silchar Subsection Conference (SILCON)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/SILCON55242.2022.10028896\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 IEEE Silchar Subsection Conference (SILCON)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/SILCON55242.2022.10028896","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Detection of Architectural Distortion using Deep Convolutional Neural Network
Breast cancer is one of the threatening diseases among women throughout the world. The early detection is the only way to cure from cancer. Architectural distortion (AD) is one of the earliest symptoms of breast cancer which is mostly malignant in nature. Computer-aided detection (CAD) and particularly deep learning (DL) gives prominent solution for the detection and diagnosis of breast cancer. This paper presents a deep convolutional neural network (DCNN) architecture designed for the automatic detection of AD in digital mammography images. The proposed deep learning based model consists of series combination of down sampler and ResNet blocks. Due to stacking of these blocks, these layers learn more complex features which help to improved in sensitivity and performance of the model. A total of 150 mammograms are considered for experimentation purpose from publicly available dataset namely, DDSM. Hence the best result obtained in the proposed approach with Leave-One-Out cross validation technique, in terms of true positive rate 86% at 0.42 false positives per image (FPs/I).