Optimizing acquisition time and injected dose in 18F-FDG PET/CT imaging using deep learning: enhancing image protocol efficiency and safety
Yulong Zeng, Weixia Chong, Zhao Ge, Jie Qin, Biyun Mo, Xingyu Mu, Wei Fu · 2026
Read Optimizing acquisition time and injected dose in 18F-FDG PET/CT imaging using deep learning: enhancing image protocol efficiency and safety by Yulong Zeng, Weixia Chong, Zhao Ge, Jie Qin, Biyun Mo, Xingyu Mu, Wei Fu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13455379.txt.
Description
Zeng and colleagues trained a deep learning network to clean up PET scans taken faster or with half the usual radioactive tracer dose. The aim was to cut patient radiation exposure and scanner time without losing detail. They studied 322 patients, faking half-dose imaging and scan times from 30 to 90 seconds per position.
A residual U-Net rebuilt the noisy images, lifting reader quality scores in the half-dose group from 3.45 at baseline to 4.33 after processing, on the five-point scale used. Numeric similarity climbed too, with structural similarity rising from 0.75 at the low end to 0.87 at the high end and peak signal-to-noise from 37.65 decibels before to 41.37 decibels after in half-dose reconstructions. Even the fastest 30-second protocol recovered strongly after processing in the reported metrics. Detection of small lesions improved by up to 16.3 percent in shortened scans, judged by two nuclear medicine physicians, supporting efficient AI-enhanced protocols.
Catalog record
| CREATOR | Yulong Zeng, Weixia Chong, Zhao Ge, Jie Qin, Biyun Mo, Xingyu Mu, Wei Fu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 42 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |