http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Automated recognition of white blood cells using deep learning
Amin Khouani,Mostafa El Habib Daho,Sidi Ahmed Mahmoudi,Mohammed Amine Chikh,Brahim Benzineb 대한의용생체공학회 2020 Biomedical Engineering Letters (BMEL) Vol.10 No.3
The detection, counting, and precise segmentation of white blood cells in cytological images are vital steps in the eff ectivediagnosis of several cancers. This paper introduces an effi cient method for automatic recognition of white blood cells inperipheral blood and bone marrow images based on deep learning to alleviate tedious tasks for hematologists in clinicalpractice. First, input image pre-processing was proposed before applying a deep neural network model adapted to cellslocalization and segmentation. Then, model outputs were improved by using combined predictions and corrections. Finally,a new algorithm that uses the cooperation between model results and spatial information was implemented to improve thesegmentation quality. To implement our model, python language, Tensorfl ow, and Keras libraries were used. The calculationswere executed using NVIDIA GPU 1080, while the datasets used in our experiments came from patients in the Hemobiologyservice of Tlemcen Hospital (Algeria). The results were promising and showed the effi ciency, power, and speed of theproposed method compared to the state-of-the-art methods. In addition to its accuracy of 95.73%, the proposed approachprovided fast predictions (less than 1 s).