Computer-Assisted Fine-Needle Aspiration Cytology of Thyroid Using Two-Stage Refined Convolutional Neural Network
Fine-needle aspiration cytology (FNAC) is regarded as one of the most important preoperative diagnostic tests for thyroid nodules. However, the traditional diagnostic process of FNAC is time-consuming, and its accuracy is highly related to the experience of the cytopathologist. Computer-aided diagno...
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-12-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2079-9292/11/24/4089 |
| _version_ | 1797460090294370304 |
|---|---|
| author | Wensi Duan Lili Gao Juan Liu Cheng Li Peng Jiang Lang Wang Hua Chen Xiaorong Sun Dehua Cao Baochuan Pang Rong Li Sai Liu |
| author_facet | Wensi Duan Lili Gao Juan Liu Cheng Li Peng Jiang Lang Wang Hua Chen Xiaorong Sun Dehua Cao Baochuan Pang Rong Li Sai Liu |
| author_sort | Wensi Duan |
| collection | DOAJ |
| description | Fine-needle aspiration cytology (FNAC) is regarded as one of the most important preoperative diagnostic tests for thyroid nodules. However, the traditional diagnostic process of FNAC is time-consuming, and its accuracy is highly related to the experience of the cytopathologist. Computer-aided diagnostic (CAD) systems are rapidly evolving to provide objective diagnostic recommendations. So far, most studies have used fixed-size patches and usually hand-select patches for model training. In this study, we develop a CAD system to address these challenges. In order to be consistent with the diagnostic working mode of cytopathologists, the system is mainly composed of two task modules: the detecting module that is responsible for detecting the regions of interest (ROIs) from the whole slide image of the FNAC, and the classification module that identifies ROIs having positive lesions. The system can then output the top-k ROIs with the highest positive probabilities for the cytopathologists to review. In order to obtain the overall good performance of the system, we compared different object detection and classification models, and used a combination of the YOLOV4 and EfficientNet networks in our system. |
| first_indexed | 2024-03-09T17:00:12Z |
| format | Article |
| id | doaj.art-0696e5807eff4b88b5a5c252d05a9848 |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-03-09T17:00:12Z |
| publishDate | 2022-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-0696e5807eff4b88b5a5c252d05a98482023-11-24T14:30:02ZengMDPI AGElectronics2079-92922022-12-011124408910.3390/electronics11244089Computer-Assisted Fine-Needle Aspiration Cytology of Thyroid Using Two-Stage Refined Convolutional Neural NetworkWensi Duan0Lili Gao1Juan Liu2Cheng Li3Peng Jiang4Lang Wang5Hua Chen6Xiaorong Sun7Dehua Cao8Baochuan Pang9Rong Li10Sai Liu11Institute of Artificial Intelligence, School of Computer Science, Wuhan University, Wuhan 430072, ChinaDepartment of Pathology, Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai 200025, ChinaInstitute of Artificial Intelligence, School of Computer Science, Wuhan University, Wuhan 430072, ChinaLanding Artificial Intelligence Center for Pathological Diagnosis, Wuhan 430070, ChinaInstitute of Artificial Intelligence, School of Computer Science, Wuhan University, Wuhan 430072, ChinaInstitute of Artificial Intelligence, School of Computer Science, Wuhan University, Wuhan 430072, ChinaInstitute of Artificial Intelligence, School of Computer Science, Wuhan University, Wuhan 430072, ChinaLanding Artificial Intelligence Center for Pathological Diagnosis, Wuhan 430070, ChinaLanding Artificial Intelligence Center for Pathological Diagnosis, Wuhan 430070, ChinaLanding Artificial Intelligence Center for Pathological Diagnosis, Wuhan 430070, ChinaLanding Artificial Intelligence Center for Pathological Diagnosis, Wuhan 430070, ChinaLanding Artificial Intelligence Center for Pathological Diagnosis, Wuhan 430070, ChinaFine-needle aspiration cytology (FNAC) is regarded as one of the most important preoperative diagnostic tests for thyroid nodules. However, the traditional diagnostic process of FNAC is time-consuming, and its accuracy is highly related to the experience of the cytopathologist. Computer-aided diagnostic (CAD) systems are rapidly evolving to provide objective diagnostic recommendations. So far, most studies have used fixed-size patches and usually hand-select patches for model training. In this study, we develop a CAD system to address these challenges. In order to be consistent with the diagnostic working mode of cytopathologists, the system is mainly composed of two task modules: the detecting module that is responsible for detecting the regions of interest (ROIs) from the whole slide image of the FNAC, and the classification module that identifies ROIs having positive lesions. The system can then output the top-k ROIs with the highest positive probabilities for the cytopathologists to review. In order to obtain the overall good performance of the system, we compared different object detection and classification models, and used a combination of the YOLOV4 and EfficientNet networks in our system.https://www.mdpi.com/2079-9292/11/24/4089two-stage CAD systemfine-needle aspiration cytology (FNAC)thyroid cytopathologydeep learningobject detection |
| spellingShingle | Wensi Duan Lili Gao Juan Liu Cheng Li Peng Jiang Lang Wang Hua Chen Xiaorong Sun Dehua Cao Baochuan Pang Rong Li Sai Liu Computer-Assisted Fine-Needle Aspiration Cytology of Thyroid Using Two-Stage Refined Convolutional Neural Network Electronics two-stage CAD system fine-needle aspiration cytology (FNAC) thyroid cytopathology deep learning object detection |
| title | Computer-Assisted Fine-Needle Aspiration Cytology of Thyroid Using Two-Stage Refined Convolutional Neural Network |
| title_full | Computer-Assisted Fine-Needle Aspiration Cytology of Thyroid Using Two-Stage Refined Convolutional Neural Network |
| title_fullStr | Computer-Assisted Fine-Needle Aspiration Cytology of Thyroid Using Two-Stage Refined Convolutional Neural Network |
| title_full_unstemmed | Computer-Assisted Fine-Needle Aspiration Cytology of Thyroid Using Two-Stage Refined Convolutional Neural Network |
| title_short | Computer-Assisted Fine-Needle Aspiration Cytology of Thyroid Using Two-Stage Refined Convolutional Neural Network |
| title_sort | computer assisted fine needle aspiration cytology of thyroid using two stage refined convolutional neural network |
| topic | two-stage CAD system fine-needle aspiration cytology (FNAC) thyroid cytopathology deep learning object detection |
| url | https://www.mdpi.com/2079-9292/11/24/4089 |
| work_keys_str_mv | AT wensiduan computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork AT liligao computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork AT juanliu computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork AT chengli computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork AT pengjiang computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork AT langwang computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork AT huachen computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork AT xiaorongsun computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork AT dehuacao computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork AT baochuanpang computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork AT rongli computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork AT sailiu computerassistedfineneedleaspirationcytologyofthyroidusingtwostagerefinedconvolutionalneuralnetwork |