PTC-MAS: A Deep Learning-Based Preoperative Automatic Assessment of Lymph Node Metastasis in Primary Thyroid Cancer
Background: Identifying cervical lymph node metastasis (LNM) in primary thyroid cancer preoperatively using ultrasound is challenging. Therefore, a non-invasive method is needed to assess LNM accurately. Purpose: To address this need, we developed the Primary Thyroid Cancer Lymph Node Metastasis Ass...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/13/10/1723 |
| _version_ | 1797600381203644416 |
|---|---|
| author | Ruqian Fu Hao Yang Dezhi Zeng Shuhan Yang Peng Luo Zhijie Yang Hua Teng Jianli Ren |
| author_facet | Ruqian Fu Hao Yang Dezhi Zeng Shuhan Yang Peng Luo Zhijie Yang Hua Teng Jianli Ren |
| author_sort | Ruqian Fu |
| collection | DOAJ |
| description | Background: Identifying cervical lymph node metastasis (LNM) in primary thyroid cancer preoperatively using ultrasound is challenging. Therefore, a non-invasive method is needed to assess LNM accurately. Purpose: To address this need, we developed the Primary Thyroid Cancer Lymph Node Metastasis Assessment System (PTC-MAS), a transfer learning-based and B-mode ultrasound images-based automatic assessment system for assessing LNM in primary thyroid cancer. Methods: The system has two parts: YOLO Thyroid Nodule Recognition System (YOLOS) for obtaining regions of interest (ROIs) of nodules, and LMM assessment system for building the LNM assessment system using transfer learning and majority voting with extracted ROIs as input. We retained the relative size features of nodules to improve the system’s performance. Results: We evaluated three transfer learning-based neural networks (DenseNet, ResNet, and GoogLeNet) and majority voting, which had the area under the curves (AUCs) of 0.802, 0.837, 0.823, and 0.858, respectively. Method III preserved relative size features and achieved higher AUCs than Method II, which fixed nodule size. YOLOS achieved high precision and sensitivity on a test set, indicating its potential for ROIs extraction. Conclusions: Our proposed PTC-MAS system effectively assesses primary thyroid cancer LNM based on preserving nodule relative size features. It has potential for guiding treatment modalities and avoiding inaccurate ultrasound results due to tracheal interference. |
| first_indexed | 2024-03-11T03:48:18Z |
| format | Article |
| id | doaj.art-0511d9194fd74716a559d0132ef6cc9f |
| institution | Directory Open Access Journal |
| issn | 2075-4418 |
| language | English |
| last_indexed | 2024-03-11T03:48:18Z |
| publishDate | 2023-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj.art-0511d9194fd74716a559d0132ef6cc9f2023-11-18T01:04:07ZengMDPI AGDiagnostics2075-44182023-05-011310172310.3390/diagnostics13101723PTC-MAS: A Deep Learning-Based Preoperative Automatic Assessment of Lymph Node Metastasis in Primary Thyroid CancerRuqian Fu0Hao Yang1Dezhi Zeng2Shuhan Yang3Peng Luo4Zhijie Yang5Hua Teng6Jianli Ren7Department of Ultrasound, The Second Affiliated Hospital of Chongqing Medical University, Chongqing 400010, ChinaDepartment of Ultrasound, The Second Affiliated Hospital of Chongqing Medical University, Chongqing 400010, ChinaDepartment of Ultrasound, The Second Affiliated Hospital of Chongqing Medical University, Chongqing 400010, ChinaMedical Data Science Academy, Chongqing Medical University, Chongqing 400010, ChinaDepartment of Ultrasound, The Second Affiliated Hospital of Chongqing Medical University, Chongqing 400010, ChinaBreast & Thyroid Surgery, The Second Affiliated Hospital of Chongqing Medical University, Chongqing 400010, ChinaDepartment of Ultrasound, The Second Affiliated Hospital of Chongqing Medical University, Chongqing 400010, ChinaDepartment of Ultrasound, The Second Affiliated Hospital of Chongqing Medical University, Chongqing 400010, ChinaBackground: Identifying cervical lymph node metastasis (LNM) in primary thyroid cancer preoperatively using ultrasound is challenging. Therefore, a non-invasive method is needed to assess LNM accurately. Purpose: To address this need, we developed the Primary Thyroid Cancer Lymph Node Metastasis Assessment System (PTC-MAS), a transfer learning-based and B-mode ultrasound images-based automatic assessment system for assessing LNM in primary thyroid cancer. Methods: The system has two parts: YOLO Thyroid Nodule Recognition System (YOLOS) for obtaining regions of interest (ROIs) of nodules, and LMM assessment system for building the LNM assessment system using transfer learning and majority voting with extracted ROIs as input. We retained the relative size features of nodules to improve the system’s performance. Results: We evaluated three transfer learning-based neural networks (DenseNet, ResNet, and GoogLeNet) and majority voting, which had the area under the curves (AUCs) of 0.802, 0.837, 0.823, and 0.858, respectively. Method III preserved relative size features and achieved higher AUCs than Method II, which fixed nodule size. YOLOS achieved high precision and sensitivity on a test set, indicating its potential for ROIs extraction. Conclusions: Our proposed PTC-MAS system effectively assesses primary thyroid cancer LNM based on preserving nodule relative size features. It has potential for guiding treatment modalities and avoiding inaccurate ultrasound results due to tracheal interference.https://www.mdpi.com/2075-4418/13/10/1723transfer learninglymph node metastasisthyroid cancerdeep learningultrasonographydiagnosis |
| spellingShingle | Ruqian Fu Hao Yang Dezhi Zeng Shuhan Yang Peng Luo Zhijie Yang Hua Teng Jianli Ren PTC-MAS: A Deep Learning-Based Preoperative Automatic Assessment of Lymph Node Metastasis in Primary Thyroid Cancer Diagnostics transfer learning lymph node metastasis thyroid cancer deep learning ultrasonography diagnosis |
| title | PTC-MAS: A Deep Learning-Based Preoperative Automatic Assessment of Lymph Node Metastasis in Primary Thyroid Cancer |
| title_full | PTC-MAS: A Deep Learning-Based Preoperative Automatic Assessment of Lymph Node Metastasis in Primary Thyroid Cancer |
| title_fullStr | PTC-MAS: A Deep Learning-Based Preoperative Automatic Assessment of Lymph Node Metastasis in Primary Thyroid Cancer |
| title_full_unstemmed | PTC-MAS: A Deep Learning-Based Preoperative Automatic Assessment of Lymph Node Metastasis in Primary Thyroid Cancer |
| title_short | PTC-MAS: A Deep Learning-Based Preoperative Automatic Assessment of Lymph Node Metastasis in Primary Thyroid Cancer |
| title_sort | ptc mas a deep learning based preoperative automatic assessment of lymph node metastasis in primary thyroid cancer |
| topic | transfer learning lymph node metastasis thyroid cancer deep learning ultrasonography diagnosis |
| url | https://www.mdpi.com/2075-4418/13/10/1723 |
| work_keys_str_mv | AT ruqianfu ptcmasadeeplearningbasedpreoperativeautomaticassessmentoflymphnodemetastasisinprimarythyroidcancer AT haoyang ptcmasadeeplearningbasedpreoperativeautomaticassessmentoflymphnodemetastasisinprimarythyroidcancer AT dezhizeng ptcmasadeeplearningbasedpreoperativeautomaticassessmentoflymphnodemetastasisinprimarythyroidcancer AT shuhanyang ptcmasadeeplearningbasedpreoperativeautomaticassessmentoflymphnodemetastasisinprimarythyroidcancer AT pengluo ptcmasadeeplearningbasedpreoperativeautomaticassessmentoflymphnodemetastasisinprimarythyroidcancer AT zhijieyang ptcmasadeeplearningbasedpreoperativeautomaticassessmentoflymphnodemetastasisinprimarythyroidcancer AT huateng ptcmasadeeplearningbasedpreoperativeautomaticassessmentoflymphnodemetastasisinprimarythyroidcancer AT jianliren ptcmasadeeplearningbasedpreoperativeautomaticassessmentoflymphnodemetastasisinprimarythyroidcancer |