Related Experiment Video
Updated: Jul 2, 2026
03:55
Computer-Aided Three-Dimensional Visualization in the Treatment of Locally Advanced Thyroid Cancer
Published on: June 9, 2023
1.1K
CT-Free Quantitative Thyroid SPECT Based on Artificial Intelligence: A Prospective Multicenter Noninferiority
Hyun Woo Chung1, Sang-Geon Cho2, Dongkyu Oh3,4
1Department of Nuclear Medicine, Konkuk University Medical Center, Konkuk University School of Medicine, Seoul, Republic of Korea.
Summary
CT-free thyroid SPECT using artificial intelligence (AI) matches conventional SPECT/CT for diagnosing thyrotoxicosis. This AI-powered method reduces radiation exposure and analysis time.
Area of Science:
- Nuclear medicine
- Radiology
- Artificial Intelligence
Background:
- Hybrid single-photon emission computed tomography (SPECT)/computed tomography (CT) is used for thyrotoxicosis diagnosis.
- Its reliance on CT and complex analysis is debated.
- This study explores CT-free SPECT with AI as an alternative.
Purpose of the Study:
- To evaluate if CT-free thyroid SPECT using AI can replace conventional SPECT/CT for thyrotoxicosis diagnosis.
- To compare diagnostic accuracy and efficiency.
- To assess radiation dose reduction.
Main Methods:
- A prospective multicenter noninferiority trial included 152 thyrotoxicosis patients.
- An AI model generated technetium thyroid uptake (TcTU) values from SPECT images without CT.
- TcTU values were compared to those from conventional SPECT/CT.
Main Results:
- CT-free SPECT showed 86.2% accuracy for Graves' disease, noninferior to SPECT/CT's 85.5%.
- Radiation exposure decreased by 33.5% (3.34 mSv to 2.22 mSv).
- CT-free SPECT analysis time was significantly reduced (4 minutes vs. 40 minutes).
Conclusions:
- CT-free SPECT with AI offers comparable diagnostic performance to conventional SPECT/CT for thyrotoxicosis.
- This AI-driven approach reduces radiation exposure and analysis time.
- It presents a viable alternative for differential diagnosis of thyrotoxicosis.
