Related Experiment Video
Updated: Nov 25, 2025
10:02
Whole-body PET/MRI of Pediatric Patients: The Details That Matter
Published on: December 19, 2017
15.0K
Reducing scan time of paediatric 99mTc-DMSA SPECT via deep learning
1Department of Nuclear Medicine, Chang Gung Memorial Hospital, No. 5, Fuxing Street, Gueishan Dist., Taoyuan, 33305, Taiwan.
Clinical Radiology
|December 19, 2020
Summary
Deep learning (DL) can significantly reduce pediatric technetium 99m (99mTc) dimercaptosuccinic acid (DMSA) SPECT scan times. This method maintains diagnostic accuracy, improving efficiency in pediatric kidney imaging.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Medicine
- Pediatric Radiology
Background:
- Technetium 99m (99mTc) dimercaptosuccinic acid (DMSA) SPECT is crucial for pediatric renal imaging.
- Long scan times can be challenging for pediatric patients.
- Deep learning (DL) offers potential solutions for optimizing imaging protocols.
Purpose of the Study:
- To assess the feasibility of reducing pediatric 99mTc-DMSA SPECT scan duration using a DL approach.
- To evaluate if shortened scan times impact diagnostic image quality and accuracy.
Main Methods:
- Retrospective analysis of 112 pediatric 99mTc-DMSA SPECT scans.
- Training a DL model on 88 scans to generate full-time SPECT images from half-time acquisitions.
- Validating the DL model's performance on the remaining 24 scans.
Main Results:
- DL-generated SPECT images from half-time acquisition demonstrated comparable image quality to standard full-time acquisition images.
- The DL-based method achieved high diagnostic performance: 91.7% accuracy, 83.3% sensitivity, and 100% specificity for detecting affected kidneys.
Conclusions:
- Deep learning shows promise in reducing pediatric 99mTc-DMSA SPECT scan times.
- This approach can maintain diagnostic accuracy, potentially improving patient comfort and workflow efficiency.
