Author: Shuo Jin; Bo Wang; Haibo Xu; Chuan Luo; Lai Wei; Wei Zhao; Xuexue Hou; Wenshuo Ma; Zhengqing Xu; Zhuozhao Zheng; Wenbo Sun; Lan Lan; Wei Zhang; Xiangdong Mu; Chenxi Shi; Zhongxiao Wang; Jihae Lee; Zijian Jin; Minggui Lin; Hongbo Jin; Liang Zhang; Jun Guo; Benqi Zhao; Zhizhong Ren; Shuhao Wang; Zheng You; Jiahong Dong; Xinghuan Wang; Jianming Wang; Wei Xu
Title: AI-assisted CT imaging analysis for COVID-19 screening: Building and deploying a medical AI system in four weeks Document date: 2020_3_23
ID: e6q92shw_11
Snippet: The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03. 19.20039354 doi: medRxiv preprint image samples from 5 different hospitals (Table S1 ) with 11 different models of CT equipments (Table S2 ) to increase the model's generalization ability. 3) We developed a three-stage annotation and quality control pipeline, allowing inexperienced data annotators to work with senior radiologists to crea.....
Document: The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03. 19.20039354 doi: medRxiv preprint image samples from 5 different hospitals (Table S1 ) with 11 different models of CT equipments (Table S2 ) to increase the model's generalization ability. 3) We developed a three-stage annotation and quality control pipeline, allowing inexperienced data annotators to work with senior radiologists to create accurate annotations, with minimal time investment from the radiologists. 4) We got reasonable training result with only 131 positive cases. Then as more data came in, we retrained the model to improve the model accuracy continuously. 5) To make the diagnosis more intuitive to the radiologists and physicians, in addition to the classification model (output positive / negative predictions), we also introduced a segmentation model that highlighted lesion regions for further examination. 6) The entire tool was delivered as a low-cost, plug-and-play device, so the hospital IT staff could set it up in a self-service way, and we could remotely upgrade the models.
Search related documents:
Co phrase search for related documents- classification model and generalization ability: 1, 2, 3
- classification model and lesion region: 1, 2
- classification model and low cost: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13
- classification model and model accuracy: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73
- classification model and model accuracy improve: 1, 2
- classification model and model generalization ability: 1
- classification model and model retrain: 1, 2
- classification model and positive case: 1, 2
- classification model and quality control: 1
- classification model and segmentation model: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26
- classification model and training result: 1
Co phrase search for related documents, hyperlinks ordered by date