Development and validation of a deep learning model for detection of breast cancers in mammography from multi-institutional datasets.
<h4>Objectives</h4>The objective of this study was to develop and validate a state-of-the-art, deep learning (DL)-based model for detecting breast cancers on mammography.<h4>Methods</h4>Mammograms in a hospital development dataset, a hospital test dataset, and a clinic test d...
| Main Authors: | , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2022-01-01
|
| Series: | PLoS ONE |
| Online Access: | https://doi.org/10.1371/journal.pone.0265751 |
| _version_ | 1811306079970131968 |
|---|---|
| author | Daiju Ueda Akira Yamamoto Naoyoshi Onoda Tsutomu Takashima Satoru Noda Shinichiro Kashiwagi Tamami Morisaki Shinya Fukumoto Masatsugu Shiba Mina Morimura Taro Shimono Ken Kageyama Hiroyuki Tatekawa Kazuki Murai Takashi Honjo Akitoshi Shimazaki Daijiro Kabata Yukio Miki |
| author_facet | Daiju Ueda Akira Yamamoto Naoyoshi Onoda Tsutomu Takashima Satoru Noda Shinichiro Kashiwagi Tamami Morisaki Shinya Fukumoto Masatsugu Shiba Mina Morimura Taro Shimono Ken Kageyama Hiroyuki Tatekawa Kazuki Murai Takashi Honjo Akitoshi Shimazaki Daijiro Kabata Yukio Miki |
| author_sort | Daiju Ueda |
| collection | DOAJ |
| description | <h4>Objectives</h4>The objective of this study was to develop and validate a state-of-the-art, deep learning (DL)-based model for detecting breast cancers on mammography.<h4>Methods</h4>Mammograms in a hospital development dataset, a hospital test dataset, and a clinic test dataset were retrospectively collected from January 2006 through December 2017 in Osaka City University Hospital and Medcity21 Clinic. The hospital development dataset and a publicly available digital database for screening mammography (DDSM) dataset were used to train and to validate the RetinaNet, one type of DL-based model, with five-fold cross-validation. The model's sensitivity and mean false positive indications per image (mFPI) and partial area under the curve (AUC) with 1.0 mFPI for both test datasets were externally assessed with the test datasets.<h4>Results</h4>The hospital development dataset, hospital test dataset, clinic test dataset, and DDSM development dataset included a total of 3179 images (1448 malignant images), 491 images (225 malignant images), 2821 images (37 malignant images), and 1457 malignant images, respectively. The proposed model detected all cancers with a 0.45-0.47 mFPI and had partial AUCs of 0.93 in both test datasets.<h4>Conclusions</h4>The DL-based model developed for this study was able to detect all breast cancers with a very low mFPI. Our DL-based model achieved the highest performance to date, which might lead to improved diagnosis for breast cancer. |
| first_indexed | 2024-04-13T08:37:59Z |
| format | Article |
| id | doaj.art-949bf8a8f4e34d038679f6f1ac589d22 |
| institution | Directory Open Access Journal |
| issn | 1932-6203 |
| language | English |
| last_indexed | 2024-04-13T08:37:59Z |
| publishDate | 2022-01-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS ONE |
| spelling | doaj.art-949bf8a8f4e34d038679f6f1ac589d222022-12-22T02:54:02ZengPublic Library of Science (PLoS)PLoS ONE1932-62032022-01-01173e026575110.1371/journal.pone.0265751Development and validation of a deep learning model for detection of breast cancers in mammography from multi-institutional datasets.Daiju UedaAkira YamamotoNaoyoshi OnodaTsutomu TakashimaSatoru NodaShinichiro KashiwagiTamami MorisakiShinya FukumotoMasatsugu ShibaMina MorimuraTaro ShimonoKen KageyamaHiroyuki TatekawaKazuki MuraiTakashi HonjoAkitoshi ShimazakiDaijiro KabataYukio Miki<h4>Objectives</h4>The objective of this study was to develop and validate a state-of-the-art, deep learning (DL)-based model for detecting breast cancers on mammography.<h4>Methods</h4>Mammograms in a hospital development dataset, a hospital test dataset, and a clinic test dataset were retrospectively collected from January 2006 through December 2017 in Osaka City University Hospital and Medcity21 Clinic. The hospital development dataset and a publicly available digital database for screening mammography (DDSM) dataset were used to train and to validate the RetinaNet, one type of DL-based model, with five-fold cross-validation. The model's sensitivity and mean false positive indications per image (mFPI) and partial area under the curve (AUC) with 1.0 mFPI for both test datasets were externally assessed with the test datasets.<h4>Results</h4>The hospital development dataset, hospital test dataset, clinic test dataset, and DDSM development dataset included a total of 3179 images (1448 malignant images), 491 images (225 malignant images), 2821 images (37 malignant images), and 1457 malignant images, respectively. The proposed model detected all cancers with a 0.45-0.47 mFPI and had partial AUCs of 0.93 in both test datasets.<h4>Conclusions</h4>The DL-based model developed for this study was able to detect all breast cancers with a very low mFPI. Our DL-based model achieved the highest performance to date, which might lead to improved diagnosis for breast cancer.https://doi.org/10.1371/journal.pone.0265751 |
| spellingShingle | Daiju Ueda Akira Yamamoto Naoyoshi Onoda Tsutomu Takashima Satoru Noda Shinichiro Kashiwagi Tamami Morisaki Shinya Fukumoto Masatsugu Shiba Mina Morimura Taro Shimono Ken Kageyama Hiroyuki Tatekawa Kazuki Murai Takashi Honjo Akitoshi Shimazaki Daijiro Kabata Yukio Miki Development and validation of a deep learning model for detection of breast cancers in mammography from multi-institutional datasets. PLoS ONE |
| title | Development and validation of a deep learning model for detection of breast cancers in mammography from multi-institutional datasets. |
| title_full | Development and validation of a deep learning model for detection of breast cancers in mammography from multi-institutional datasets. |
| title_fullStr | Development and validation of a deep learning model for detection of breast cancers in mammography from multi-institutional datasets. |
| title_full_unstemmed | Development and validation of a deep learning model for detection of breast cancers in mammography from multi-institutional datasets. |
| title_short | Development and validation of a deep learning model for detection of breast cancers in mammography from multi-institutional datasets. |
| title_sort | development and validation of a deep learning model for detection of breast cancers in mammography from multi institutional datasets |
| url | https://doi.org/10.1371/journal.pone.0265751 |
| work_keys_str_mv | AT daijuueda developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT akirayamamoto developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT naoyoshionoda developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT tsutomutakashima developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT satorunoda developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT shinichirokashiwagi developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT tamamimorisaki developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT shinyafukumoto developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT masatsugushiba developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT minamorimura developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT taroshimono developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT kenkageyama developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT hiroyukitatekawa developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT kazukimurai developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT takashihonjo developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT akitoshishimazaki developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT daijirokabata developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets AT yukiomiki developmentandvalidationofadeeplearningmodelfordetectionofbreastcancersinmammographyfrommultiinstitutionaldatasets |