Classification of MR-Detected Additional Lesions in Patients With Breast Cancer Using a Combination of Radiomics Analysis and Machine Learning
ObjectiveThis study was conducted in order to investigate the feasibility of using radiomics analysis (RA) with machine learning algorithms based on breast magnetic resonance (MR) images for discriminating malignant from benign MR-detected additional lesions in patients with primary breast cancer.Ma...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2021-12-01
|
| Series: | Frontiers in Oncology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fonc.2021.744460/full |
| _version_ | 1818572924094775296 |
|---|---|
| author | Hyo-jae Lee Anh-Tien Nguyen So Yeon Ki Jong Eun Lee Luu-Ngoc Do Min Ho Park Min Ho Park Ji Shin Lee Ji Shin Lee Hye Jung Kim Ilwoo Park Ilwoo Park Ilwoo Park Hyo Soon Lim Hyo Soon Lim |
| author_facet | Hyo-jae Lee Anh-Tien Nguyen So Yeon Ki Jong Eun Lee Luu-Ngoc Do Min Ho Park Min Ho Park Ji Shin Lee Ji Shin Lee Hye Jung Kim Ilwoo Park Ilwoo Park Ilwoo Park Hyo Soon Lim Hyo Soon Lim |
| author_sort | Hyo-jae Lee |
| collection | DOAJ |
| description | ObjectiveThis study was conducted in order to investigate the feasibility of using radiomics analysis (RA) with machine learning algorithms based on breast magnetic resonance (MR) images for discriminating malignant from benign MR-detected additional lesions in patients with primary breast cancer.Materials and MethodsOne hundred seventy-four MR-detected additional lesions (benign, n = 86; malignancy, n = 88) from 158 patients with ipsilateral primary breast cancer from a tertiary medical center were included in this retrospective study. The entire data were randomly split to training (80%) and independent test sets (20%). In addition, 25 patients (benign, n = 21; malignancy, n = 15) from another tertiary medical center were included for the external test. Radiomics features that were extracted from three regions-of-interest (ROIs; intratumor, peritumor, combined) using fat-saturated T1-weighted images obtained by subtracting pre- from postcontrast images (SUB) and T2-weighted image (T2) were utilized to train the support vector machine for the binary classification. A decision tree method was utilized to build a classifier model using clinical imaging interpretation (CII) features assessed by radiologists. Area under the receiver operating characteristic curve (AUROC), accuracy, sensitivity, and specificity were used to compare the diagnostic performance.ResultsThe RA models trained using radiomics features from the intratumor-ROI showed comparable performance to the CII model (accuracy, AUROC: 73.3%, 69.6% for the SUB RA model; 70.0%, 75.1% for the T2 RA model; 73.3%, 72.0% for the CII model). The diagnostic performance increased when the radiomics and CII features were combined to build a fusion model. The fusion model that combines the CII features and radiomics features from multiparametric MRI data demonstrated the highest performance with an accuracy of 86.7% and an AUROC of 91.1%. The external test showed a similar pattern where the fusion models demonstrated higher levels of performance compared with the RA- or CII-only models. The accuracy and AUROC of the SUB+T2 RA+CII model in the external test were 80.6% and 91.4%, respectively.ConclusionOur study demonstrated the feasibility of using RA with machine learning approach based on multiparametric MRI for quantitatively characterizing MR-detected additional lesions. The fusion model demonstrated an improved diagnostic performance over the models trained with either RA or CII alone. |
| first_indexed | 2024-12-15T00:04:08Z |
| format | Article |
| id | doaj.art-39ea7ab626464e2f96a3b96afcd71066 |
| institution | Directory Open Access Journal |
| issn | 2234-943X |
| language | English |
| last_indexed | 2024-12-15T00:04:08Z |
| publishDate | 2021-12-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Oncology |
| spelling | doaj.art-39ea7ab626464e2f96a3b96afcd710662022-12-21T22:42:48ZengFrontiers Media S.A.Frontiers in Oncology2234-943X2021-12-011110.3389/fonc.2021.744460744460Classification of MR-Detected Additional Lesions in Patients With Breast Cancer Using a Combination of Radiomics Analysis and Machine LearningHyo-jae Lee0Anh-Tien Nguyen1So Yeon Ki2Jong Eun Lee3Luu-Ngoc Do4Min Ho Park5Min Ho Park6Ji Shin Lee7Ji Shin Lee8Hye Jung Kim9Ilwoo Park10Ilwoo Park11Ilwoo Park12Hyo Soon Lim13Hyo Soon Lim14Department of Radiology, Chonnam National University Hospital, Gwangju, South KoreaDepartment of Radiology, Chonnam National University Hospital, Gwangju, South KoreaDepartment of Radiology, Chonnam National University Hwasun Hospital, Hwasun-gun, South KoreaDepartment of Radiology, Chonnam National University Hospital, Gwangju, South KoreaDepartment of Radiology, Chonnam National University, Gwangju, South KoreaDepartment of Radiology, Chonnam National University, Gwangju, South KoreaDepartment of Surgery, Chonnam National University Hwasun Hospital, Hwasun-gun, South KoreaDepartment of Radiology, Chonnam National University, Gwangju, South KoreaDepartment of Pathology, Chonnam National University Hwasun Hospital, Hwasun-gun, South KoreaDepartment of Radiology, School of Medicine, Kyungpook National University, Kyungpook National University Chilgok Hospital, Daegu, South KoreaDepartment of Radiology, Chonnam National University Hospital, Gwangju, South KoreaDepartment of Radiology, Chonnam National University, Gwangju, South KoreaDepartment of Artificial Intelligence Convergence, Chonnam National University, Gwangju, South KoreaDepartment of Radiology, Chonnam National University Hwasun Hospital, Hwasun-gun, South KoreaDepartment of Radiology, Chonnam National University, Gwangju, South KoreaObjectiveThis study was conducted in order to investigate the feasibility of using radiomics analysis (RA) with machine learning algorithms based on breast magnetic resonance (MR) images for discriminating malignant from benign MR-detected additional lesions in patients with primary breast cancer.Materials and MethodsOne hundred seventy-four MR-detected additional lesions (benign, n = 86; malignancy, n = 88) from 158 patients with ipsilateral primary breast cancer from a tertiary medical center were included in this retrospective study. The entire data were randomly split to training (80%) and independent test sets (20%). In addition, 25 patients (benign, n = 21; malignancy, n = 15) from another tertiary medical center were included for the external test. Radiomics features that were extracted from three regions-of-interest (ROIs; intratumor, peritumor, combined) using fat-saturated T1-weighted images obtained by subtracting pre- from postcontrast images (SUB) and T2-weighted image (T2) were utilized to train the support vector machine for the binary classification. A decision tree method was utilized to build a classifier model using clinical imaging interpretation (CII) features assessed by radiologists. Area under the receiver operating characteristic curve (AUROC), accuracy, sensitivity, and specificity were used to compare the diagnostic performance.ResultsThe RA models trained using radiomics features from the intratumor-ROI showed comparable performance to the CII model (accuracy, AUROC: 73.3%, 69.6% for the SUB RA model; 70.0%, 75.1% for the T2 RA model; 73.3%, 72.0% for the CII model). The diagnostic performance increased when the radiomics and CII features were combined to build a fusion model. The fusion model that combines the CII features and radiomics features from multiparametric MRI data demonstrated the highest performance with an accuracy of 86.7% and an AUROC of 91.1%. The external test showed a similar pattern where the fusion models demonstrated higher levels of performance compared with the RA- or CII-only models. The accuracy and AUROC of the SUB+T2 RA+CII model in the external test were 80.6% and 91.4%, respectively.ConclusionOur study demonstrated the feasibility of using RA with machine learning approach based on multiparametric MRI for quantitatively characterizing MR-detected additional lesions. The fusion model demonstrated an improved diagnostic performance over the models trained with either RA or CII alone.https://www.frontiersin.org/articles/10.3389/fonc.2021.744460/fullbreast neoplasmsmagnetic resonance imagingmachine learningradiomicsultrasonography |
| spellingShingle | Hyo-jae Lee Anh-Tien Nguyen So Yeon Ki Jong Eun Lee Luu-Ngoc Do Min Ho Park Min Ho Park Ji Shin Lee Ji Shin Lee Hye Jung Kim Ilwoo Park Ilwoo Park Ilwoo Park Hyo Soon Lim Hyo Soon Lim Classification of MR-Detected Additional Lesions in Patients With Breast Cancer Using a Combination of Radiomics Analysis and Machine Learning Frontiers in Oncology breast neoplasms magnetic resonance imaging machine learning radiomics ultrasonography |
| title | Classification of MR-Detected Additional Lesions in Patients With Breast Cancer Using a Combination of Radiomics Analysis and Machine Learning |
| title_full | Classification of MR-Detected Additional Lesions in Patients With Breast Cancer Using a Combination of Radiomics Analysis and Machine Learning |
| title_fullStr | Classification of MR-Detected Additional Lesions in Patients With Breast Cancer Using a Combination of Radiomics Analysis and Machine Learning |
| title_full_unstemmed | Classification of MR-Detected Additional Lesions in Patients With Breast Cancer Using a Combination of Radiomics Analysis and Machine Learning |
| title_short | Classification of MR-Detected Additional Lesions in Patients With Breast Cancer Using a Combination of Radiomics Analysis and Machine Learning |
| title_sort | classification of mr detected additional lesions in patients with breast cancer using a combination of radiomics analysis and machine learning |
| topic | breast neoplasms magnetic resonance imaging machine learning radiomics ultrasonography |
| url | https://www.frontiersin.org/articles/10.3389/fonc.2021.744460/full |
| work_keys_str_mv | AT hyojaelee classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT anhtiennguyen classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT soyeonki classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT jongeunlee classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT luungocdo classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT minhopark classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT minhopark classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT jishinlee classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT jishinlee classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT hyejungkim classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT ilwoopark classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT ilwoopark classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT ilwoopark classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT hyosoonlim classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning AT hyosoonlim classificationofmrdetectedadditionallesionsinpatientswithbreastcancerusingacombinationofradiomicsanalysisandmachinelearning |