DeepAMO: a multi-slice, multi-view anthropomorphic model observer for visual detection tasks performed on volume images

X-ray Microtomography to Assess Determinants of In Vivo N-Butyl Cyanoacrylate Glubran®2 Polymerization: A Rabbit-Model Study

Although introduced decades ago, few cyanoacrylate glues have been approved for endovascular use, despite evidence of their usefulness, notably for complex procedures suchas hemostatic embolization. Indications include massive bleeding requiring emergent hemostasis and prevention of severe bleeding during scheduled surgery to remove a hypervascular tumor. Adding radiopaque Lipiodol Ultra Fluid® (LUF) modulates glue polymerization and allows fluoroscopic guidance, but few comparative in vivo studies have assessed the impact of the resulting change in glue concentration or of other factors such as target-vessel blood flow. In a rabbit model, we used ex vivo X-ray microtomography to assess the results of in vivo renal-artery embolization by various mixtures of N-butyl cyanoacrylate (NBCA), metacryloxysulfolane, and LUF. Overall, penetration to the superficial interlobular arteries was achieved in about two-thirds of cases and into the capillaries in nearly half the cases, while cast fragmentation was seen in slightly more than half the cases. Greater NBCA dilution and the blocked-blood-flow technique were independently associated with greater distality of penetration. Blocked-blood-flow injection was independently associated with absence of fragmentation, capillary penetration, a shorter cast-to-capsule distance, and higher cast attenuation. A larger mixture volume was independently associated with higher indexed cast ratio and deeper penetration. Finally, microtomography is an adapted tool to assess ex vivo distribution of glue cast.

Improving the detectability of overactive glands in dual-phase parathyroid SPECT/CT: a Monte Carlo simulation study

Objective. SPECT-CT is a standard procedure conducted before minimally invasive surgery for the treatment of primary hyperthyroidism. In order to improve image quality, it is important to know how defect detectability is influenced by acquisition and processing parameters. The objective of this study is to continue prior physical phantom optimization studies by performing Monte Carlo simulations for the dual phase parathyroid SPECT-CT protocol using a digital anthropomorphic phantom.Methods. The dual phase parathyroid SPECT-CT imaging procedure with 99mTc-Sestamibi was simulated using the previously extensively validated SIMIND software for the first time. An anthropomorphic ZUBAL based phantom was built to represent an adenoma. Its diameter was set to 0.76 cm which corresponded to more than three times the pixel size and the target-to-background ratio was set to 16:1 based on previous studies. Four different collimators were tested. Contrast-to-noise (CNR) values were determined for different scatter correction options and processing parameter values. The OSEM algorithm was used for image reconstruction.Results. CNR values were improved from about zero (LEGP collimator, 16 iterations, attenuation correction: on, scatter correction: off) up to 3.7 (LEUHR collimator, 16 iterations, attenuation correction: on, scatter correction: off). The subjective visual assessment of detectability on simulated images agreed with the quantitative CNR values.Conclusion. Higher resolution collimators gave better CNR as confirmed by similar studies. The effect of scatter correction was found beneficial only if both the resolution and sensitivity of the collimator were relatively high. This is a significant finding since there is a shortage of definitive guideline on the use of scatter correction for parathyroid SPECT imaging.