Yasunori Nagayama1, Shota Tanoue2, Taihei Inoue2, Seitaro Oda2, Takeshi Nakaura2, Daisuke Utsunomiya3, Yasuyuki Yamashita2. 1. Department of Diagnostic Radiology, Graduate School of Medical Sciences, Kumamoto University, 1-1-1, Honjo, Chuo-ku, Kumamoto, 860-8556, Japan. y.nagayama1980@gmail.com. 2. Department of Diagnostic Radiology, Graduate School of Medical Sciences, Kumamoto University, 1-1-1, Honjo, Chuo-ku, Kumamoto, 860-8556, Japan. 3. Department of Diagnostic Radiology, Yokohama City University Hospital, 3-9 Fukuura, Kanazawa-ku, Yokohama, 236-0004, Japan.
Abstract
OBJECTIVES: To evaluate the image quality and optimal energies of virtual monoenergetic images (VMIs) from dual-layer spectral detector computed tomography (DLCT) in multiphasic pancreatic CT and investigate whether low-keV VMI at the portal venous phase (PVP) provides sufficient tumor conspicuity and arterial depiction relative to conventional pancreatic parenchymal phase (PPP) images. METHODS: Forty-eight patients with pancreatic ductal adenocarcinoma (PDAC) underwent contrast-enhanced DLCT during PPP and PVP. Conventional polyenergetic images (PEIs) and VMI at 40-100 keV (VMI40-100, 10-keV increments) were reconstructed at each enhancement phase. Image noise and the contrast-to-noise ratio (CNR) of the pancreas, tumors, arteries, and veins were quantified. Two radiologists independently assessed tumor conspicuity, margin delineation, image noise, sharpness of pancreatic duct, and depiction of arteries and veins on a five-point scale. Size-specific dose estimate (SSDE) was calculated. RESULTS: Image noise for VMI40-100 was significantly lower than that for PEI (p < 0.01). The CNR in VMI increased gradually with decreasing energy; CNRs for VMI40-60 were significantly greater than that for PEI (p < 0.01). All subjective VMI scores were maximized at VMI40, followed by VMI50-60, all of which were significantly better than of PEI (p < 0.01). Objective and subjective image qualities of VMI40-50 at the PVP were equivalent to or even better compared with conventional PPP images. No significant difference in SSDE was observed between phases (p = 0.10). CONCLUSIONS: DLCT-VMI improved the subjective and objective image quality in multiphasic pancreatic CT for patients with PDAC. Low-keV PVP imaging may yield diagnostically adequate tumor conspicuity and arterial assessment compared with polyenergetic PPP images. KEY POINTS: • Low-keV VMI from DLCT yields better subjective and objective image quality of multiphasic pancreas CT in comparison with conventional PEI for the assessment of pancreatic ductal adenocarcinoma. • Tumor conspicuity and depiction of peripancreatic vasculature were maximized at VMI 40without an increase in the image noise. • Low-keV VMI of the portal venous phase provides sufficient tumor conspicuity and arterial depiction, potentially allowing the early detection and local staging of PDAC on routine abdominal CT performed for various clinical indications.
OBJECTIVES: To evaluate the image quality and optimal energies of virtual monoenergetic images (VMIs) from dual-layer spectral detector computed tomography (DLCT) in multiphasic pancreatic CT and investigate whether low-keV VMI at the portal venous phase (PVP) provides sufficient tumor conspicuity and arterial depiction relative to conventional pancreatic parenchymal phase (PPP) images. METHODS: Forty-eight patients with pancreatic ductal adenocarcinoma (PDAC) underwent contrast-enhanced DLCT during PPP and PVP. Conventional polyenergetic images (PEIs) and VMI at 40-100 keV (VMI40-100, 10-keV increments) were reconstructed at each enhancement phase. Image noise and the contrast-to-noise ratio (CNR) of the pancreas, tumors, arteries, and veins were quantified. Two radiologists independently assessed tumor conspicuity, margin delineation, image noise, sharpness of pancreatic duct, and depiction of arteries and veins on a five-point scale. Size-specific dose estimate (SSDE) was calculated. RESULTS: Image noise for VMI40-100 was significantly lower than that for PEI (p < 0.01). The CNR in VMI increased gradually with decreasing energy; CNRs for VMI40-60 were significantly greater than that for PEI (p < 0.01). All subjective VMI scores were maximized at VMI40, followed by VMI50-60, all of which were significantly better than of PEI (p < 0.01). Objective and subjective image qualities of VMI40-50 at the PVP were equivalent to or even better compared with conventional PPP images. No significant difference in SSDE was observed between phases (p = 0.10). CONCLUSIONS: DLCT-VMI improved the subjective and objective image quality in multiphasic pancreatic CT for patients with PDAC. Low-keV PVP imaging may yield diagnostically adequate tumor conspicuity and arterial assessment compared with polyenergetic PPP images. KEY POINTS: • Low-keV VMI from DLCT yields better subjective and objective image quality of multiphasic pancreas CT in comparison with conventional PEI for the assessment of pancreatic ductal adenocarcinoma. • Tumor conspicuity and depiction of peripancreatic vasculature were maximized at VMI 40without an increase in the image noise. • Low-keV VMI of the portal venous phase provides sufficient tumor conspicuity and arterial depiction, potentially allowing the early detection and local staging of PDAC on routine abdominal CT performed for various clinical indications.
Authors: Mahmoud M Al-Hawary; Isaac R Francis; Suresh T Chari; Elliot K Fishman; David M Hough; David S Lu; Michael Macari; Alec J Megibow; Frank H Miller; Koenraad J Mortele; Nipun B Merchant; Rebecca M Minter; Eric P Tamm; Dushyant V Sahani; Diane M Simeone Journal: Radiology Date: 2014-01 Impact factor: 11.105
Authors: Joel G Fletcher; Maurits J Wiersema; Michael A Farrell; Jeff L Fidler; Lawrence J Burgart; Takashi Koyama; C Daniel Johnson; David H Stephens; Ellen M Ward; W Scott Harmsen Journal: Radiology Date: 2003-10 Impact factor: 11.105
Authors: Rupert W Prokesch; Lawrence C Chow; Christopher F Beaulieu; Roland Bammer; R Brooke Jeffrey Journal: Radiology Date: 2002-09 Impact factor: 11.105
Authors: C Mattevi; J Garnier; U Marchese; J Ewald; M Gilabert; F Poizat; G Piana; J R Delpero; O Turrini Journal: BMC Surg Date: 2020-08-05 Impact factor: 2.102