Deep Learning-Based Pathology Image Analysis Enhances Magee Feature Correlation With Oncotype DX Breast Recurrence Score
BackgroundOncotype DX Recurrence Score (RS) has been widely used to predict chemotherapy benefits in patients with estrogen receptor-positive breast cancer. Studies showed that the features used in Magee equations correlate with RS. We aimed to examine whether deep learning (DL)-based histology imag...
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-06-01
|
| Series: | Frontiers in Medicine |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fmed.2022.886763/full |
| _version_ | 1811257876284440576 |
|---|---|
| author | Hongxiao Li Hongxiao Li Jigang Wang Jigang Wang Zaibo Li Melad Dababneh Fusheng Wang Peng Zhao Geoffrey H. Smith George Teodoro Meijie Li Jun Kong Jun Kong Jun Kong Xiaoxian Li |
| author_facet | Hongxiao Li Hongxiao Li Jigang Wang Jigang Wang Zaibo Li Melad Dababneh Fusheng Wang Peng Zhao Geoffrey H. Smith George Teodoro Meijie Li Jun Kong Jun Kong Jun Kong Xiaoxian Li |
| author_sort | Hongxiao Li |
| collection | DOAJ |
| description | BackgroundOncotype DX Recurrence Score (RS) has been widely used to predict chemotherapy benefits in patients with estrogen receptor-positive breast cancer. Studies showed that the features used in Magee equations correlate with RS. We aimed to examine whether deep learning (DL)-based histology image analyses can enhance such correlations.MethodsWe retrieved 382 cases with RS diagnosed between 2011 and 2015 from the Emory University and the Ohio State University. All patients received surgery. DL models were developed to detect nuclei of tumor cells and tumor-infiltrating lymphocytes (TILs) and segment tumor cell nuclei in hematoxylin and eosin (H&E) stained histopathology whole slide images (WSIs). Based on the DL-based analysis, we derived image features from WSIs, such as tumor cell number, TIL number variance, and nuclear grades. The entire patient cohorts were divided into one training set (125 cases) and two validation sets (82 and 175 cases) based on the data sources and WSI resolutions. The training set was used to train the linear regression models to predict RS. For prediction performance comparison, we used independent variables from Magee features alone or the combination of WSI-derived image and Magee features.ResultsThe Pearson’s correlation coefficients between the actual RS and predicted RS by DL-based analysis were 0.7058 (p-value = 1.32 × 10–13) and 0.5041 (p-value = 1.15 × 10–12) for the validation sets 1 and 2, respectively. The adjusted R2 values using Magee features alone are 0.3442 and 0.2167 in the two validation sets, respectively. In contrast, the adjusted R2 values were enhanced to 0.4431 and 0.2182 when WSI-derived imaging features were jointly used with Magee features.ConclusionOur results suggest that DL-based digital pathological features can enhance Magee feature correlation with RS. |
| first_indexed | 2024-04-12T18:04:28Z |
| format | Article |
| id | doaj.art-8338778c420f4163820873aed515ec86 |
| institution | Directory Open Access Journal |
| issn | 2296-858X |
| language | English |
| last_indexed | 2024-04-12T18:04:28Z |
| publishDate | 2022-06-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Medicine |
| spelling | doaj.art-8338778c420f4163820873aed515ec862022-12-22T03:22:02ZengFrontiers Media S.A.Frontiers in Medicine2296-858X2022-06-01910.3389/fmed.2022.886763886763Deep Learning-Based Pathology Image Analysis Enhances Magee Feature Correlation With Oncotype DX Breast Recurrence ScoreHongxiao Li0Hongxiao Li1Jigang Wang2Jigang Wang3Zaibo Li4Melad Dababneh5Fusheng Wang6Peng Zhao7Geoffrey H. Smith8George Teodoro9Meijie Li10Jun Kong11Jun Kong12Jun Kong13Xiaoxian Li14Department of Mathematics and Statistics, Georgia State University, Atlanta, GA, United StatesInstitute of Biomedical Engineering, Chinese Academy of Medical Sciences and Peking Union Medical College, Tianjin, ChinaDepartment of Pathology, The Affiliated Hospital of Qingdao University, Qingdao, ChinaDepartment of Pathology and Laboratory Medicine, Emory University, Atlanta, GA, United StatesDepartment of Pathology, The Ohio State University, Columbus, OH, United StatesDepartment of Pathology and Laboratory Medicine, Emory University, Atlanta, GA, United StatesDepartment of Computer Science, Stony Brook University, Stony Brook, NY, United StatesDepartment of Pathology, The Affiliated Hospital of Qingdao University, Qingdao, ChinaDepartment of Pathology and Laboratory Medicine, Emory University, Atlanta, GA, United StatesDepartment of Computer Science, Federal University of Minas Gerais, Belo Horizonte, BrazilDepartment of Mathematics and Statistics, Georgia State University, Atlanta, GA, United StatesDepartment of Mathematics and Statistics, Georgia State University, Atlanta, GA, United StatesDepartment of Computer Science, Georgia State University, Atlanta, GA, United StatesDepartment of Computer Science, Emory University, Atlanta, GA, United StatesDepartment of Pathology and Laboratory Medicine, Emory University, Atlanta, GA, United StatesBackgroundOncotype DX Recurrence Score (RS) has been widely used to predict chemotherapy benefits in patients with estrogen receptor-positive breast cancer. Studies showed that the features used in Magee equations correlate with RS. We aimed to examine whether deep learning (DL)-based histology image analyses can enhance such correlations.MethodsWe retrieved 382 cases with RS diagnosed between 2011 and 2015 from the Emory University and the Ohio State University. All patients received surgery. DL models were developed to detect nuclei of tumor cells and tumor-infiltrating lymphocytes (TILs) and segment tumor cell nuclei in hematoxylin and eosin (H&E) stained histopathology whole slide images (WSIs). Based on the DL-based analysis, we derived image features from WSIs, such as tumor cell number, TIL number variance, and nuclear grades. The entire patient cohorts were divided into one training set (125 cases) and two validation sets (82 and 175 cases) based on the data sources and WSI resolutions. The training set was used to train the linear regression models to predict RS. For prediction performance comparison, we used independent variables from Magee features alone or the combination of WSI-derived image and Magee features.ResultsThe Pearson’s correlation coefficients between the actual RS and predicted RS by DL-based analysis were 0.7058 (p-value = 1.32 × 10–13) and 0.5041 (p-value = 1.15 × 10–12) for the validation sets 1 and 2, respectively. The adjusted R2 values using Magee features alone are 0.3442 and 0.2167 in the two validation sets, respectively. In contrast, the adjusted R2 values were enhanced to 0.4431 and 0.2182 when WSI-derived imaging features were jointly used with Magee features.ConclusionOur results suggest that DL-based digital pathological features can enhance Magee feature correlation with RS.https://www.frontiersin.org/articles/10.3389/fmed.2022.886763/fulldeep learning-based algorithmdigital pathologyOncotype DX scoreER+ breast cancerMagee equation |
| spellingShingle | Hongxiao Li Hongxiao Li Jigang Wang Jigang Wang Zaibo Li Melad Dababneh Fusheng Wang Peng Zhao Geoffrey H. Smith George Teodoro Meijie Li Jun Kong Jun Kong Jun Kong Xiaoxian Li Deep Learning-Based Pathology Image Analysis Enhances Magee Feature Correlation With Oncotype DX Breast Recurrence Score Frontiers in Medicine deep learning-based algorithm digital pathology Oncotype DX score ER+ breast cancer Magee equation |
| title | Deep Learning-Based Pathology Image Analysis Enhances Magee Feature Correlation With Oncotype DX Breast Recurrence Score |
| title_full | Deep Learning-Based Pathology Image Analysis Enhances Magee Feature Correlation With Oncotype DX Breast Recurrence Score |
| title_fullStr | Deep Learning-Based Pathology Image Analysis Enhances Magee Feature Correlation With Oncotype DX Breast Recurrence Score |
| title_full_unstemmed | Deep Learning-Based Pathology Image Analysis Enhances Magee Feature Correlation With Oncotype DX Breast Recurrence Score |
| title_short | Deep Learning-Based Pathology Image Analysis Enhances Magee Feature Correlation With Oncotype DX Breast Recurrence Score |
| title_sort | deep learning based pathology image analysis enhances magee feature correlation with oncotype dx breast recurrence score |
| topic | deep learning-based algorithm digital pathology Oncotype DX score ER+ breast cancer Magee equation |
| url | https://www.frontiersin.org/articles/10.3389/fmed.2022.886763/full |
| work_keys_str_mv | AT hongxiaoli deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT hongxiaoli deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT jigangwang deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT jigangwang deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT zaiboli deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT meladdababneh deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT fushengwang deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT pengzhao deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT geoffreyhsmith deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT georgeteodoro deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT meijieli deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT junkong deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT junkong deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT junkong deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore AT xiaoxianli deeplearningbasedpathologyimageanalysisenhancesmageefeaturecorrelationwithoncotypedxbreastrecurrencescore |