Ensemble of fine‐tuned convolutional neural networks for urine sediment microscopic image classification
In this study, an ensemble of fine‐tuned convolutional neural networks (CNNs) is proposed. As CNN training requires large annotated data, which are lacking in the field of urine sediment microscopic image processing, the authors first pre‐trained the CNNs, including ResNet50 and GoogLeNet, and devel...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2020-02-01
|
| Series: | IET Computer Vision |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/iet-cvi.2018.5829 |
| _version_ | 1797684821419360256 |
|---|---|
| author | Wenqian Liu Weihong Li Weiguo Gong |
| author_facet | Wenqian Liu Weihong Li Weiguo Gong |
| author_sort | Wenqian Liu |
| collection | DOAJ |
| description | In this study, an ensemble of fine‐tuned convolutional neural networks (CNNs) is proposed. As CNN training requires large annotated data, which are lacking in the field of urine sediment microscopic image processing, the authors first pre‐trained the CNNs, including ResNet50 and GoogLeNet, and developed AlexNet on an ImageNet dataset. Thereafter, some of the weights of the pre‐trained CNNs were transferred to the urine sediment microscopic image dataset. To guide fine‐tuning of the learning rate and cascading features, the hierarchical nature of features in different convolutional layers was investigated by visualising the CNN. Then, they combined three CNNs as an ensemble of CNNs to decrease the differences and impurity interference among features of urine sediment microscopic image. These fusion features were employed to train the fully connected neural network for classification. In this study, they improved the accuracy of each CNN by an average of 2.2% through fine‐tuning of the learning rate and cascading features. Moreover, the better experimental results were achieved compared with other state‐of‐the‐art methods and indicated that a 97% classification accuracy can be attained. |
| first_indexed | 2024-03-12T00:35:16Z |
| format | Article |
| id | doaj.art-8fc6d7e94c4744dc9cf8858859e24abf |
| institution | Directory Open Access Journal |
| issn | 1751-9632 1751-9640 |
| language | English |
| last_indexed | 2024-03-12T00:35:16Z |
| publishDate | 2020-02-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Computer Vision |
| spelling | doaj.art-8fc6d7e94c4744dc9cf8858859e24abf2023-09-15T09:56:19ZengWileyIET Computer Vision1751-96321751-96402020-02-01141182510.1049/iet-cvi.2018.5829Ensemble of fine‐tuned convolutional neural networks for urine sediment microscopic image classificationWenqian Liu0Weihong Li1Weiguo Gong2Key Lab of Optoelectronic Technology and Systems Ministry of EducationCollege of Optoelectronic Engineering, Chongqing UniversityRoom 1303, Main Building, No.174 Shazheng Street, Shapingba DistrictChongqingPeople's Republic of ChinaKey Lab of Optoelectronic Technology and Systems Ministry of EducationCollege of Optoelectronic Engineering, Chongqing UniversityRoom 1303, Main Building, No.174 Shazheng Street, Shapingba DistrictChongqingPeople's Republic of ChinaKey Lab of Optoelectronic Technology and Systems Ministry of EducationCollege of Optoelectronic Engineering, Chongqing UniversityRoom 1303, Main Building, No.174 Shazheng Street, Shapingba DistrictChongqingPeople's Republic of ChinaIn this study, an ensemble of fine‐tuned convolutional neural networks (CNNs) is proposed. As CNN training requires large annotated data, which are lacking in the field of urine sediment microscopic image processing, the authors first pre‐trained the CNNs, including ResNet50 and GoogLeNet, and developed AlexNet on an ImageNet dataset. Thereafter, some of the weights of the pre‐trained CNNs were transferred to the urine sediment microscopic image dataset. To guide fine‐tuning of the learning rate and cascading features, the hierarchical nature of features in different convolutional layers was investigated by visualising the CNN. Then, they combined three CNNs as an ensemble of CNNs to decrease the differences and impurity interference among features of urine sediment microscopic image. These fusion features were employed to train the fully connected neural network for classification. In this study, they improved the accuracy of each CNN by an average of 2.2% through fine‐tuning of the learning rate and cascading features. Moreover, the better experimental results were achieved compared with other state‐of‐the‐art methods and indicated that a 97% classification accuracy can be attained.https://doi.org/10.1049/iet-cvi.2018.5829urine sediment microscopic image processingpre-trained CNNsurine sediment microscopic image datasetlearning ratecascading featuresconvolutional layers |
| spellingShingle | Wenqian Liu Weihong Li Weiguo Gong Ensemble of fine‐tuned convolutional neural networks for urine sediment microscopic image classification IET Computer Vision urine sediment microscopic image processing pre-trained CNNs urine sediment microscopic image dataset learning rate cascading features convolutional layers |
| title | Ensemble of fine‐tuned convolutional neural networks for urine sediment microscopic image classification |
| title_full | Ensemble of fine‐tuned convolutional neural networks for urine sediment microscopic image classification |
| title_fullStr | Ensemble of fine‐tuned convolutional neural networks for urine sediment microscopic image classification |
| title_full_unstemmed | Ensemble of fine‐tuned convolutional neural networks for urine sediment microscopic image classification |
| title_short | Ensemble of fine‐tuned convolutional neural networks for urine sediment microscopic image classification |
| title_sort | ensemble of fine tuned convolutional neural networks for urine sediment microscopic image classification |
| topic | urine sediment microscopic image processing pre-trained CNNs urine sediment microscopic image dataset learning rate cascading features convolutional layers |
| url | https://doi.org/10.1049/iet-cvi.2018.5829 |
| work_keys_str_mv | AT wenqianliu ensembleoffinetunedconvolutionalneuralnetworksforurinesedimentmicroscopicimageclassification AT weihongli ensembleoffinetunedconvolutionalneuralnetworksforurinesedimentmicroscopicimageclassification AT weiguogong ensembleoffinetunedconvolutionalneuralnetworksforurinesedimentmicroscopicimageclassification |