An Approach for Training Data Enrichment and Batch Labeling in AI+MRI Aided Diagnosis
Training data enrichment is a key factor in artificial intelligence (AI) technology development. At present, the bottleneck problem is that the quantity and type of labeled training data in valid samples are unable to meet the requirements of AI+MRI aided diagnosis. In this paper, an effective appro...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Science Press
2018-12-01
|
| Series: | Chinese Journal of Magnetic Resonance |
| Subjects: | |
| Online Access: | http://html.rhhz.net/bpxzz/html/20180405.htm |
| _version_ | 1818293920016105472 |
|---|---|
| author | WANG Hong-zhi ZHAO Di YANG Li-qin XIA Tian ZHOU Xiao-yue MIAO Zhi-ying |
| author_facet | WANG Hong-zhi ZHAO Di YANG Li-qin XIA Tian ZHOU Xiao-yue MIAO Zhi-ying |
| author_sort | WANG Hong-zhi |
| collection | DOAJ |
| description | Training data enrichment is a key factor in artificial intelligence (AI) technology development. At present, the bottleneck problem is that the quantity and type of labeled training data in valid samples are unable to meet the requirements of AI+MRI aided diagnosis. In this paper, an effective approach to solve the problem was presented. High resolution isotropic multi-dimensional data of regions of interests from patients or healthy volunteers were first acquired via a series of scanning on clinical MRI scanners, including quantitative T1, T2, proton density (Pd) and apparent diffusion coefficient (ADC) measurements. These data were then used as the ground truth, from which different types of images associated with different imaging sequences and parameters were obtained with a virtual MRI technology. The type of the images with the best boundary resolution were then selected manually by experienced doctors, on which three-dimensional mask matrix was obtained by manual contouring and labeling, serving as the template for other types of images. This enrichment method was developed as a software platform, which could provide sufficient quantity of image data from a small number of positive cases, thus meeting the data training enrichment requirement of AI+MRI diagnosis at low cost and with high efficiency. |
| first_indexed | 2024-12-13T03:23:31Z |
| format | Article |
| id | doaj.art-a378a6d8e4b6492598b4449974982aeb |
| institution | Directory Open Access Journal |
| issn | 1000-4556 1000-4556 |
| language | zho |
| last_indexed | 2024-12-13T03:23:31Z |
| publishDate | 2018-12-01 |
| publisher | Science Press |
| record_format | Article |
| series | Chinese Journal of Magnetic Resonance |
| spelling | doaj.art-a378a6d8e4b6492598b4449974982aeb2022-12-22T00:01:19ZzhoScience PressChinese Journal of Magnetic Resonance1000-45561000-45562018-12-0135444745610.11938/cjmr20182658An Approach for Training Data Enrichment and Batch Labeling in AI+MRI Aided DiagnosisWANG Hong-zhi0ZHAO Di1YANG Li-qin2XIA Tian3ZHOU Xiao-yue4MIAO Zhi-ying5Shanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai 200062, ChinaInstitute of Computing Technology, Chinese Academy of Sciences, Beijing 100190, ChinaHuashan Hospital, Fudan University, Shanghai 200040, ChinaShanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai 200062, China Siemens Healthineers, Shanghai 201318, ChinaSchool of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai 200093, ChinaTraining data enrichment is a key factor in artificial intelligence (AI) technology development. At present, the bottleneck problem is that the quantity and type of labeled training data in valid samples are unable to meet the requirements of AI+MRI aided diagnosis. In this paper, an effective approach to solve the problem was presented. High resolution isotropic multi-dimensional data of regions of interests from patients or healthy volunteers were first acquired via a series of scanning on clinical MRI scanners, including quantitative T1, T2, proton density (Pd) and apparent diffusion coefficient (ADC) measurements. These data were then used as the ground truth, from which different types of images associated with different imaging sequences and parameters were obtained with a virtual MRI technology. The type of the images with the best boundary resolution were then selected manually by experienced doctors, on which three-dimensional mask matrix was obtained by manual contouring and labeling, serving as the template for other types of images. This enrichment method was developed as a software platform, which could provide sufficient quantity of image data from a small number of positive cases, thus meeting the data training enrichment requirement of AI+MRI diagnosis at low cost and with high efficiency.http://html.rhhz.net/bpxzz/html/20180405.htmartificial intelligence (AI)magnetic resonance imaging (MRI)training data enrichmentbatch labelingimage aided diagnosis |
| spellingShingle | WANG Hong-zhi ZHAO Di YANG Li-qin XIA Tian ZHOU Xiao-yue MIAO Zhi-ying An Approach for Training Data Enrichment and Batch Labeling in AI+MRI Aided Diagnosis Chinese Journal of Magnetic Resonance artificial intelligence (AI) magnetic resonance imaging (MRI) training data enrichment batch labeling image aided diagnosis |
| title | An Approach for Training Data Enrichment and Batch Labeling in AI+MRI Aided Diagnosis |
| title_full | An Approach for Training Data Enrichment and Batch Labeling in AI+MRI Aided Diagnosis |
| title_fullStr | An Approach for Training Data Enrichment and Batch Labeling in AI+MRI Aided Diagnosis |
| title_full_unstemmed | An Approach for Training Data Enrichment and Batch Labeling in AI+MRI Aided Diagnosis |
| title_short | An Approach for Training Data Enrichment and Batch Labeling in AI+MRI Aided Diagnosis |
| title_sort | approach for training data enrichment and batch labeling in ai mri aided diagnosis |
| topic | artificial intelligence (AI) magnetic resonance imaging (MRI) training data enrichment batch labeling image aided diagnosis |
| url | http://html.rhhz.net/bpxzz/html/20180405.htm |
| work_keys_str_mv | AT wanghongzhi anapproachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis AT zhaodi anapproachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis AT yangliqin anapproachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis AT xiatian anapproachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis AT zhouxiaoyue anapproachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis AT miaozhiying anapproachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis AT wanghongzhi approachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis AT zhaodi approachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis AT yangliqin approachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis AT xiatian approachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis AT zhouxiaoyue approachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis AT miaozhiying approachfortrainingdataenrichmentandbatchlabelinginaimriaideddiagnosis |