Deep Learning-Enabled Clinically Applicable CT Planbox for Stroke With High Accuracy and Repeatability
BackgroundComputed tomography (CT) plays an essential role in classifying stroke, quantifying penumbra size and supporting stroke-relevant radiomics studies. However, it is difficult to acquire standard, accurate and repeatable images during follow-up. Therefore, we invented an intelligent CT to eva...
| Main Authors: | , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-03-01
|
| Series: | Frontiers in Neurology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fneur.2022.755492/full |
| _version_ | 1817986282704338944 |
|---|---|
| author | Yang Wang Junkai Zhu Jinli Zhao Wenyi Li Xin Zhang Xiaolin Meng Taige Chen Ming Li Meiping Ye Renfang Hu Shidan Dou Huayin Hao Xiaofen Zhao Xiaoming Wu Wei Hu Cheng Li Xiaole Fan Liyun Jiang Xiaofan Lu Fangrong Yan |
| author_facet | Yang Wang Junkai Zhu Jinli Zhao Wenyi Li Xin Zhang Xiaolin Meng Taige Chen Ming Li Meiping Ye Renfang Hu Shidan Dou Huayin Hao Xiaofen Zhao Xiaoming Wu Wei Hu Cheng Li Xiaole Fan Liyun Jiang Xiaofan Lu Fangrong Yan |
| author_sort | Yang Wang |
| collection | DOAJ |
| description | BackgroundComputed tomography (CT) plays an essential role in classifying stroke, quantifying penumbra size and supporting stroke-relevant radiomics studies. However, it is difficult to acquire standard, accurate and repeatable images during follow-up. Therefore, we invented an intelligent CT to evaluate stroke during the entire follow-up.MethodsWe deployed a region proposal network (RPN) and V-Net to endow traditional CT with intelligence. Specifically, facial detection was accomplished by identifying adjacent jaw positions through training and testing an RPN on 76,382 human faces using a preinstalled 2-dimensional camera; two regions of interest (ROIs) were segmented by V-Net on another training set with 295 subjects, and the moving distance of scanning couch was calculated based on a pre-generated calibration table. Multiple cohorts including 1,124 patients were used for performance validation under three clinical scenarios.ResultsCranial Automatic Planbox Imaging Towards AmeLiorating neuroscience (CAPITAL)-CT was invented. RPN model had an error distance of 4.46 ± 0.02 pixels with a success rate of 98.7% in the training set and 100% with 2.23 ± 0.10 pixels in the testing set. V-Net-derived segmentation maintained a clinically tolerable distance error, within 3 mm on average, and all lines presented with a tolerable angle error, within 3° on average in all boundaries. Real-time, accurate, and repeatable automatic scanning was accomplished with and a lower radiation exposure dose (all P < 0.001).ConclusionsCAPITAL-CT generated standard and reproducible images that could simplify the work of radiologists, which would be of great help in the follow-up of stroke patients and in multifield research in neuroscience. |
| first_indexed | 2024-04-14T00:06:59Z |
| format | Article |
| id | doaj.art-a3172565efd84e4f976502c40280a7fb |
| institution | Directory Open Access Journal |
| issn | 1664-2295 |
| language | English |
| last_indexed | 2024-04-14T00:06:59Z |
| publishDate | 2022-03-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neurology |
| spelling | doaj.art-a3172565efd84e4f976502c40280a7fb2022-12-22T02:23:29ZengFrontiers Media S.A.Frontiers in Neurology1664-22952022-03-011310.3389/fneur.2022.755492755492Deep Learning-Enabled Clinically Applicable CT Planbox for Stroke With High Accuracy and RepeatabilityYang Wang0Junkai Zhu1Jinli Zhao2Wenyi Li3Xin Zhang4Xiaolin Meng5Taige Chen6Ming Li7Meiping Ye8Renfang Hu9Shidan Dou10Huayin Hao11Xiaofen Zhao12Xiaoming Wu13Wei Hu14Cheng Li15Xiaole Fan16Liyun Jiang17Xiaofan Lu18Fangrong Yan19Department of Radiology, Zhujiang Hospital, Southern Medical University, Guangzhou, ChinaState Key Laboratory of Natural Medicines, Research Center of Biostatistics and Computational Pharmacy, China Pharmaceutical University, Nanjing, ChinaDepartment of Radiology, The Affiliated Hospital of Nantong University, Nantong, ChinaDepartment of Endocrinology, Tongren Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, ChinaDepartment of Neurology, Drum Tower Hospital, Medical School and The State Key Laboratory of Pharmaceutical Biotechnology, Institute of Brain Science, Nanjing University, Nanjing, ChinaResearch & Advanced Algorithm Department of HSW BU, Shanghai United Imaging Healthcare Co., Ltd., Shanghai, ChinaMedical School of Nanjing University, Nanjing, ChinaDepartment of Radiology, Zhujiang Hospital, Southern Medical University, Guangzhou, ChinaDepartment of Radiology, Zhujiang Hospital, Southern Medical University, Guangzhou, ChinaCalibration Physical Algorithm Department of CT BU, Shanghai United Imaging Healthcare Co., Ltd., Shanghai, ChinaResearch & Advanced Algorithm Department of HSW BU, Shanghai United Imaging Healthcare Co., Ltd., Shanghai, ChinaResearch & Advanced Algorithm Department of HSW BU, Shanghai United Imaging Healthcare Co., Ltd., Shanghai, ChinaClinical Workflow and Clinical Verification Department of CT BU, Shanghai United Imaging Healthcare Co., Ltd., Shanghai, ChinaClinical Workflow and Clinical Verification Department of CT BU, Shanghai United Imaging Healthcare Co., Ltd., Shanghai, China0Department of CT BU, Shanghai United Imaging Healthcare Co., Ltd., Shanghai, ChinaResearch & Advanced Algorithm Department of HSW BU, Shanghai United Imaging Healthcare Co., Ltd., Shanghai, China1Department of Radiology, The Second Affiliated Hospital of Nantong University, Nantong, ChinaState Key Laboratory of Natural Medicines, Research Center of Biostatistics and Computational Pharmacy, China Pharmaceutical University, Nanjing, ChinaState Key Laboratory of Natural Medicines, Research Center of Biostatistics and Computational Pharmacy, China Pharmaceutical University, Nanjing, ChinaState Key Laboratory of Natural Medicines, Research Center of Biostatistics and Computational Pharmacy, China Pharmaceutical University, Nanjing, ChinaBackgroundComputed tomography (CT) plays an essential role in classifying stroke, quantifying penumbra size and supporting stroke-relevant radiomics studies. However, it is difficult to acquire standard, accurate and repeatable images during follow-up. Therefore, we invented an intelligent CT to evaluate stroke during the entire follow-up.MethodsWe deployed a region proposal network (RPN) and V-Net to endow traditional CT with intelligence. Specifically, facial detection was accomplished by identifying adjacent jaw positions through training and testing an RPN on 76,382 human faces using a preinstalled 2-dimensional camera; two regions of interest (ROIs) were segmented by V-Net on another training set with 295 subjects, and the moving distance of scanning couch was calculated based on a pre-generated calibration table. Multiple cohorts including 1,124 patients were used for performance validation under three clinical scenarios.ResultsCranial Automatic Planbox Imaging Towards AmeLiorating neuroscience (CAPITAL)-CT was invented. RPN model had an error distance of 4.46 ± 0.02 pixels with a success rate of 98.7% in the training set and 100% with 2.23 ± 0.10 pixels in the testing set. V-Net-derived segmentation maintained a clinically tolerable distance error, within 3 mm on average, and all lines presented with a tolerable angle error, within 3° on average in all boundaries. Real-time, accurate, and repeatable automatic scanning was accomplished with and a lower radiation exposure dose (all P < 0.001).ConclusionsCAPITAL-CT generated standard and reproducible images that could simplify the work of radiologists, which would be of great help in the follow-up of stroke patients and in multifield research in neuroscience.https://www.frontiersin.org/articles/10.3389/fneur.2022.755492/fullstrokedeep learningcomputed tomographyautomatic cranial scanningaccurate and repeatable images |
| spellingShingle | Yang Wang Junkai Zhu Jinli Zhao Wenyi Li Xin Zhang Xiaolin Meng Taige Chen Ming Li Meiping Ye Renfang Hu Shidan Dou Huayin Hao Xiaofen Zhao Xiaoming Wu Wei Hu Cheng Li Xiaole Fan Liyun Jiang Xiaofan Lu Fangrong Yan Deep Learning-Enabled Clinically Applicable CT Planbox for Stroke With High Accuracy and Repeatability Frontiers in Neurology stroke deep learning computed tomography automatic cranial scanning accurate and repeatable images |
| title | Deep Learning-Enabled Clinically Applicable CT Planbox for Stroke With High Accuracy and Repeatability |
| title_full | Deep Learning-Enabled Clinically Applicable CT Planbox for Stroke With High Accuracy and Repeatability |
| title_fullStr | Deep Learning-Enabled Clinically Applicable CT Planbox for Stroke With High Accuracy and Repeatability |
| title_full_unstemmed | Deep Learning-Enabled Clinically Applicable CT Planbox for Stroke With High Accuracy and Repeatability |
| title_short | Deep Learning-Enabled Clinically Applicable CT Planbox for Stroke With High Accuracy and Repeatability |
| title_sort | deep learning enabled clinically applicable ct planbox for stroke with high accuracy and repeatability |
| topic | stroke deep learning computed tomography automatic cranial scanning accurate and repeatable images |
| url | https://www.frontiersin.org/articles/10.3389/fneur.2022.755492/full |
| work_keys_str_mv | AT yangwang deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT junkaizhu deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT jinlizhao deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT wenyili deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT xinzhang deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT xiaolinmeng deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT taigechen deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT mingli deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT meipingye deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT renfanghu deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT shidandou deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT huayinhao deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT xiaofenzhao deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT xiaomingwu deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT weihu deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT chengli deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT xiaolefan deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT liyunjiang deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT xiaofanlu deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability AT fangrongyan deeplearningenabledclinicallyapplicablectplanboxforstrokewithhighaccuracyandrepeatability |