-
Bernstein Washington posted an update 6 months, 1 week ago
It takes on average of 7.5 s to generate a complete sCT (88 slices) for a patient on our GPU. The average MAE in HU between the sCT and actual patient CT (within the body contour) is 48.5 ± 6 HU with our method. The maximum dose difference to the target is 1.3%. This study demonstrates that an sCT can be generated in a multicentric context, with fewer pre-processing steps while being fast and accurate. © 2020 Institute of Physics and Engineering in Medicine.Glycosylated hemoglobin (HbA1c) is an important criterion for the diagnosis of diabetes and indicator of blood glucose level. But the red blood cell (RBC) lifespan heterogeneity is sufficient to influence on HbA1c interpretation. In this study, we recruited 115 patients with diabetes mellitus and 81 nondiabetic controls. The HbA1c and the RBC lifespan was detected by high performance liquid chromatography and the advanced CO breath detection method, respectively. Potential correlations of gender and age with HbA1c were analyzed and a receiver operator characteristic (ROC) curve was generated to get the HbA1c cut-off for every RBC lifespan group. It was confirmed that HbA1c has no correlation with gender and age. And the correlation formula between HbA1c diagnostic criteria and RBC lifespan was derived to correct the HbA1c diagnostic criteria using the least-square method. The RBC-lifespan-corrected HbA1c diagnostic criteria provided 100% sensitivity and specificity for the diagnosis of diabetes in the experimental set and was not refuted in the validated set. The diagnostic value of HbA1c is positively correlated with the RBC lifespan, and 4 patients with hyperglycemia, whose HbA1c values are lower than the general diagnosis criterion 6.5%, were still considered to be diabetic according to this formula, that is, the application of this formula may help us to eliminate 2.2% misdiagnosis rate of the current diagnostic criteria. To provide more accurate detection results, the effect of RBC lifespan is necessary to be taken into account when HbA1c is used as a clinical indicator. © 2020 IOP Publishing Ltd.The thermal transport of monolayer MoS2, grown by chemical vapor deposition (CVD) method, was studied in this work. A novel approach was developed to transfer monolayer MoS2 onto suspended microelectrothermal system device, where a nano-manipulator in a scanning electron microscope was employed to accomplish the feat. This nano-manipulator-assisted transferring gives a high sample yield with relatively good sample quality compared to the traditional wet/dry transfer methods. Temperature-dependent thermal conductivity of monolayer MoS2 was measured by suspended-pads thermal bridge technique, with thermal conductivity value slightly lower than the exfoliated samples due to the phonon-defects scattering for CVD grown samples. Further extension of the current transfer method was demonstrated on few-layer graphite, where suspended graphite flakes that were free of surface ripples and with high thermal conductance were shown.We developed polymeric scaffolds that can provide both topographical and electrical stimuli on mouse neural stem cells (mNSCs) for potential use in nerve tissue engineering. In contrast to conventional patterning techniques such as imprinting, soft/photolithography, and three-dimensional printing, microgroove patterns were generated by using aligned electrospun fibers as templates, via a process denoted as electrospun fiber-template lithography (EFTL). The preparation of polyvinylpyrrolidone (PVP) fibers, followed by the deposition of poly(lactic-co-glycolic acid) (PLGA) and the removal of the fiber template, produced freestanding PLGA scaffolds with microgrooves having widths of 1.72 ± 0.24 µm. The subsequent deposition of polypyrrole (PPy) via chemical oxidative polymerization added conductivity to the microgrooved PLGA scaffolds. The resultant scaffolds were cytocompatible with mNSCs. The microgroove patterns enhanced the alignment and elongation of mNSCs, and the PPy layer promoted the interaction of cells with the surface by forming more and longer filopodia compared with the nonconductive surface. Finally, the neuron differentiation of mNSCs was evaluated by monitoring the Tuj-1 neuronal gene expression. The presence of both microgrooves and the conductive PPy layer enhanced the neuronal differentiation of mNSCs even without electrical stimulation, and the neuronal differentiation was further enhanced by stimulation with a sufficient electrical pulse (1.0 V). © 2020 IOP Publishing Ltd.The strongly magnetostrictive TbFe2 compound has been epitaxially grown on Z-cut Lithium Niobate (LiNbO3) substrates after the deposition of various buffer layers (Mo, Ti and Ti/Mo). Detailed and combined RHEED and x-ray analysis permitted to unravel the in-plane and relative orientation relationships (OR) of the different materials in the system. Despite the use of different templates with different structural orders, similar final OR are eventually found between the piezoelectric substrate and the magnetic layer. The structural and magnetic properties are analyzed in order to get a TbFe2 layer of optimum quality to build a magnetostrictive/piezoelectric hybrid system with efficient strain mediated coupling. Such systems are of interest for the development of magnetic sensors as well as for the electric control of magnetization.We present DeepDose, a deep learning framework for fast dose calculations in radiation therapy. Given a patient anatomy and linear-accelerator IMRT multi-leaf-collimator shape or segment, a novel set of physics-based inputs is calculated that encode the linac machine parameters into the underlying anatomy. These inputs are then used to train a deep convolutional network to derive the dose distribution of individual MLC shapes on a given patient anatomy. In this work we demonstrate the proof-of-concept application of DeepDose on 101 prostate patients treated in our clinic with fixed-beam IMRT. The ground-truth data used for training, validation and testing of the prediction were calculated with a state-of-the-art MonteCarlo dose engine at 1% statistical uncertainty per segment. A deep convolution network was trained using the data of 80 patients at the clinically used 3 mm3grid spacing while 10 patients were used for validation. CL316243 nmr For another 11 independent test patients, the network was able to accurately estimate the segment doses from the clinical plans of each patient passing the clinical QA when compared with the MonteCarlo calculations, yielding on average 99.