Photon-Counting CT Scan Phantom Study: Stability of Radiomics Features DOI Creative Commons
Lama Dawi,

Kodjodenis Amouzouga,

Serge Muller

et al.

Diagnostics, Journal Year: 2025, Volume and Issue: 15(6), P. 649 - 649

Published: March 7, 2025

Background/Objectives: To evaluate and optimize the reconstruction parameters of images acquired with a photon-counting CT scanner to achieve stable radiomics signal. Methods: Radiomics is quantitative imaging biomarker correlated survival in oncology patients. Implementing clinical routine remains challenging due feature’s instability. Photon-counting scans use innovative technology directly converting photons into electrical signals resulting higher-resolution reduced artifacts. This study used two organic phantoms: natural wet sponge dry sausage. UHR were using NAEOTOM Alpha scan (Siemens) 0.4 mm slice thickness 0.3 × pixel size. Tube current voltage fixed at 112 mA 120 KvP. A total 24 parameter sets obtained by combining different values kernel (Br), iterative (QIR), spectral (keV), matrix Ten successive acquisitions on both phantoms. In total, 93 radiomic features extracted an ROI default Pyradiomic 3.0.1. Each stability was evaluated coefficient variation (CV) within each set. Results: Of sets, 5 selected based best image quality seven radiologists three radiology technologists. considered set when CV less than 15%. Feature impacted phantom used. The most combination included 90 65 tested sausage respectively. It configured Br36, QIR 4, 60 keV, 1024 Conclusions: Images offer promising feature optimal configurations that could be applied setting.

Language: Английский

Photon-Counting CT Scan Phantom Study: Stability of Radiomics Features DOI Creative Commons
Lama Dawi,

Kodjodenis Amouzouga,

Serge Muller

et al.

Diagnostics, Journal Year: 2025, Volume and Issue: 15(6), P. 649 - 649

Published: March 7, 2025

Background/Objectives: To evaluate and optimize the reconstruction parameters of images acquired with a photon-counting CT scanner to achieve stable radiomics signal. Methods: Radiomics is quantitative imaging biomarker correlated survival in oncology patients. Implementing clinical routine remains challenging due feature’s instability. Photon-counting scans use innovative technology directly converting photons into electrical signals resulting higher-resolution reduced artifacts. This study used two organic phantoms: natural wet sponge dry sausage. UHR were using NAEOTOM Alpha scan (Siemens) 0.4 mm slice thickness 0.3 × pixel size. Tube current voltage fixed at 112 mA 120 KvP. A total 24 parameter sets obtained by combining different values kernel (Br), iterative (QIR), spectral (keV), matrix Ten successive acquisitions on both phantoms. In total, 93 radiomic features extracted an ROI default Pyradiomic 3.0.1. Each stability was evaluated coefficient variation (CV) within each set. Results: Of sets, 5 selected based best image quality seven radiologists three radiology technologists. considered set when CV less than 15%. Feature impacted phantom used. The most combination included 90 65 tested sausage respectively. It configured Br36, QIR 4, 60 keV, 1024 Conclusions: Images offer promising feature optimal configurations that could be applied setting.

Language: Английский

Citations

0