Thyroid Imaging Programme

Artificial Intelligence in Ultrasound Thyroid Imaging

Our radiologists and radiographers are proud to partner with Endocrinology Department, Dr Chng Chiaw Ling, a Nurturing Clinician Researcher Scheme (NCRS) awardee, and data scientists from Data Science Artificial Intelligence Laboratory (DSAIL) in applying machine learning algorithms to thyroid ultrasound imaging. These efforts will support improved nodule characterization, risk stratification, and clinical decision-making. 

Contact 

Please contact for more information. 

• Dr Kheok Si Wei - kheok.si.wei@singhealth.com.sg 
• Ms Wong Yuet Wah - wong.yuet.wah@sgh.com.sg 
• Ms Kho Ying Ying -  kho.ying.ying@sgh.com.sg
• Ms Catherine Tan Wan Ting - catherine.tan.w.t@sgh.com.sg
• Ms Yew Jia Shuen - yew.jia.shuen@sgh.com.sg 
• Prof Chan Ling Ling -  chan.ling.ling@singhealth.com.sg

Factors Affecting Adherence to Fine Needle Aspiration Recommendations of TI-RADS 4 Thyroid Nodules

ACR TI-RADS recommends standardization of ultrasound reports of thyroid nodules by categorizing malignancy risk and providing management recommendations. Fine needle aspiration (FNA) thyroid nodules demonstrating ultrasound features suspicious of malignancy is recommended. However, adherence to TI-RADS recommendations is highly variable and literature sparse in this space. Dr Melissa Lee worked with a multinational team to delve into the complex interplay of factors affecting adherence to TI-RADS guidelines.

 Contact

 Please contact for more information. 

• Dr Melissa Lee Shuhui - melissa.lee.s.h@singhealth.com.sg

Imaging Recurrent Disease in Thyroid Cancer Post-Thyroidectomy

Widespread use of imaging has helped the detection of thyroid cancer. Most of these cancer remain indolent; but a patient subset requires surgical treatment and may experience recurrent disease in the cervical nodes, thyroid bed, and distant lung and bony metastases. Ultrasound (US) is the primary modality used in evaluation of recurrent disease, but CT, MRI, radioactive iodine imaging, and PET/CT can also aid in accurate diagnosis and characterization of recurrent disease. Radiologists play a central role in the detection of recurrent tumors and in guiding treatment in these patients by having a thorough understanding of the utility of these multimodal imaging techniques and imaging characteristics of recurrent disease.  

Contact 

Please contact for more information. 

• Dr Chua Wei Ming - chua.wei.ming@singhealth.com.sg
• Dr Charlene Tang Yu Lin - charlene.tang.y.l@singhealth.com.sg
• Dr Kelvin Loke Siu Hoong - kelvin.loke.s.h@singhealth.com.sg
• Dr Robert Chen Chun - robert.chen@singhealth.com.sg

Publications and Presentation:

1. Lee J, Lee MS, Wiggins R, Jridi A, Anzai Y. Factors Affecting Adherence to Fine Needle Aspiration Recommendations of TI-RADS 4 Thyroid Nodules. J Am Coll Radiol. 2025 Apr;22(4):444-448. doi: 10.1016/j.jacr.2024.10.013. Epub 2024 Oct 28.
2. Chua WM, Tang CYL, Loke KSH, Lam WW, Yang SP, Lee MS, Hou W, Lim MYS, Lim KC, Chen RC. Differentiated Thyroid Cancer after Thyroidectomy. Radiographics. 2024 Oct;44(10):e240021. doi: 10.1148/rg.240021. PMID: 39235963.
3. Chng CL, Xiang Y, Kheok SW, Kristen AL, Lim AYY, Tian KJY, Wong YW, Koh HS, Faranurshella BO, Le Q, Tan JH. Thyroid Nodule Segmentation in Ultrasound Images Using Transformer Models with Masked Autoencoder Pre-training. IMAGINE AI 2024.