Scholarlink

Purpose <p>Long-axial field-of-view PET scanners capture multi-organ tracer distribution with high sensitivity, enabling lower dose dynamic protocols and dual-tracer imaging for comprehensive disease characterization. However, reducing dose may compromise data quality and time-activity curve (TAC) fitting, leading to higher bias in kinetic parameters. Parametric imaging poses further challenges due to noise amplification in voxel-based modelling. We explore the potential of deep learning denoising (DL-DN) to improve quantification for low-dose dynamic PET.</p> Methods <p>Using 16 [<sup>18</sup>F]FDG PET studies from the PennPET Explorer, we trained a DL framework on 10-min images from late-phase uptake (static data) that were sub-sampled from 1/2 to 1/300 of the counts. This model was used to denoise early-to-late dynamic frame images. Its impact on quantification was evaluated using compartmental modelling and voxel-based graphical analysis for parametric imaging for single- and dual-tracer dynamic studies with [<sup>18</sup>F]FDG and [<sup>18</sup>F]FGln at original (injected) and reduced (sub-sampled) doses. Quantification differences were evaluated for the area under the curve of TACs, K<sub>i</sub> for [<sup>18</sup>F]FDG and V<sub>T</sub> for [<sup>18</sup>F]FGln, and parametric images.</p> Results <p>DL-DN consistently improved image quality across all dynamic frames, systematically enhancing TAC consistency and reducing tissue-dependent bias and variability in K<sub>i</sub> and V<sub>T</sub> down to 40 MBq doses. DL-DN preserved tumor heterogeneity in Logan V<sub>T</sub> images and delineation of high-flux regions in Patlak K<sub>i</sub> maps. In a /[<sup>18</sup>F]FDG dual-tracer study, bias trends aligned with single-tracer results but showed reduced accuracy for [¹⁸F]FGln in breast lesions at very low doses (4 MBq).</p> Conclusion <p>This study demonstrates that applying DL-DN trained on static [<sup>18</sup>F]FDG PET images to dynamic [<sup>18</sup>F]FDG and [<sup>18</sup>F]FGln PET can permit significantly reduced doses, preserving accurate FDG K<sub>i</sub> and FGln V<sub>T</sub> measurements, and enhancing parametric image quality. DL-DN shows promise for improving dynamic PET quantification at reduced doses, including novel dual-tracer studies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Impact of deep learning denoising on kinetic modelling for low-dose dynamic PET: application to single- and dual-tracer imaging protocols

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年3月12日

查看全文
  • Florence M. Muller,
  • Elizabeth J. Li,
  • Margaret E. Daube-Witherspoon,
  • Austin R. Pantel,
  • Corinde E. Wiers,
  • Jacob G. Dubroff,
  • Christian Vanhove,
  • Stefaan Vandenberghe,
  • Joel S. Karp

摘要

Purpose

Long-axial field-of-view PET scanners capture multi-organ tracer distribution with high sensitivity, enabling lower dose dynamic protocols and dual-tracer imaging for comprehensive disease characterization. However, reducing dose may compromise data quality and time-activity curve (TAC) fitting, leading to higher bias in kinetic parameters. Parametric imaging poses further challenges due to noise amplification in voxel-based modelling. We explore the potential of deep learning denoising (DL-DN) to improve quantification for low-dose dynamic PET.

Methods

Using 16 [18F]FDG PET studies from the PennPET Explorer, we trained a DL framework on 10-min images from late-phase uptake (static data) that were sub-sampled from 1/2 to 1/300 of the counts. This model was used to denoise early-to-late dynamic frame images. Its impact on quantification was evaluated using compartmental modelling and voxel-based graphical analysis for parametric imaging for single- and dual-tracer dynamic studies with [18F]FDG and [18F]FGln at original (injected) and reduced (sub-sampled) doses. Quantification differences were evaluated for the area under the curve of TACs, Ki for [18F]FDG and VT for [18F]FGln, and parametric images.

Results

DL-DN consistently improved image quality across all dynamic frames, systematically enhancing TAC consistency and reducing tissue-dependent bias and variability in Ki and VT down to 40 MBq doses. DL-DN preserved tumor heterogeneity in Logan VT images and delineation of high-flux regions in Patlak Ki maps. In a /[18F]FDG dual-tracer study, bias trends aligned with single-tracer results but showed reduced accuracy for [¹⁸F]FGln in breast lesions at very low doses (4 MBq).

Conclusion

This study demonstrates that applying DL-DN trained on static [18F]FDG PET images to dynamic [18F]FDG and [18F]FGln PET can permit significantly reduced doses, preserving accurate FDG Ki and FGln VT measurements, and enhancing parametric image quality. DL-DN shows promise for improving dynamic PET quantification at reduced doses, including novel dual-tracer studies.