Automated Detection and Segmentation of Bone Metastases on Spine MRI Using U-Net: A Multicenter Study.

OBJECTIVE: To develop and evaluate a deep learning model for automated segmentation and detection of bone metastasis on spinal MRI. MATERIALS AND METHODS: We included whole spine MRI scans of adult patients with bone metastasis: 662 MRI series from 302 patients (63.5 ± 11.5 years; male:female, 151:151) from three study centers obtained between January 2015 and August 2021 for training and internal testing (random split into 536 and 126 series, respectively) and 49 MRI series from 20 patients (65.9 ± 11.5 years; male:female, 11:9) from another center obtained between January 2018 and August 2020 for external testing. Three sagittal MRI sequences, including non-contrast T1-weighted image (T1), contrast-enhanced T1-weighted Dixon fat-only image (FO), and contrast-enhanced fat-suppressed T1-weighted image (CE), were used. Seven models trained using the 2D and 3D U-Nets were developed with different combinations (T1, FO, CE, T1 + FO, T1 + CE, FO + CE, and T1 + FO + CE). The segmentation performance was evaluated using Dice coefficient, pixel-wise recall, and pixel-wise precision. The detection performance was analyzed using per-lesion sensitivity and a free-response receiver operating characteristic curve. The performance of the model was compared with that of five radiologists using the external test set. RESULTS: The 2D U-Net T1 + CE model exhibited superior segmentation performance in the external test compared to the other models, with a Dice coefficient of 0.699 and pixel-wise recall of 0.653. The T1 + CE model achieved per-lesion sensitivities of 0.828 (497/600) and 0.857 (150/175) for metastases in the internal and external tests, respectively. The radiologists demonstrated a mean per-lesion sensitivity of 0.746 and a mean per-lesion positive predictive value of 0.701 in the external test. CONCLUSION: The deep learning models proposed for automated segmentation and detection of bone metastases on spinal MRI demonstrated high diagnostic performance.

ZnO nanowire-based nonvolatile memory devices with Al2O3 layers as storage nodes.

Top-gate ZnO nanowire field-effect transistors (FETs) with Al2O3 gate dielectric layers as storage nodes were fabricated and their memory effects were characterized in this work. The Al2O3 layers deposited on the ZnO nanowire channels were utilized not only as gate dielectric ones but also as charge trapping ones. For a representative top-gate ZnO nanowire FET, its I(DS)-V(GS) characteristics for the double sweep of the gate voltages exhibit the counterclockwise hysteresis and the threshold voltage shift. The gate voltage in the pulse form was applied for 1 s, and the threshold voltage shift of I(DS)-V(GS) characteristics was extended from 0.3 to 0.8 V compared with that for the double sweep. In this ZnO nanowire FET, negative charge carriers originated from the gate electrode are stored in the Al2O3 layer for applied negative gate voltages, and they are extracted for applied positive gate voltages.

Silicon nanowire-based tunneling field-effect transistors on flexible plastic substrates.

A technique to implement silicon nanowire (SiNW)-based tunneling field-effect transistors (TFETs) on flexible plastic substrates is developed for the first time. The p-i-n configured Si NWs are obtained from an Si wafer using a conventional top-down CMOS-compatible technology, and they are then transferred onto the plastic substrate. Based on gate-controlled band-to-band tunneling (BTBT) as their working principle, the SiNW-based TFETs show normal p-channel switching behavior with a threshold voltage of -1.86 V and a subthreshold swing of 827 mV/dec. In addition, ambipolar conduction is observed due to the presence of the BTBT between the heavily doped p+ drain and n+ channel regions, indicating that our TFETs can operate in the n-channel mode as well. Furthermore, the BTBT generation rates for both the p-channel and n-channel operating modes are nearly independent of the bending state (strain = 0.8%) of the plastic substrate.

ZnO nanowire field-effect transistors with floating gate nodes of Au nanoparticles.

The memory characteristics of top-gate single ZnO nanowire-based field-effect transistors (FETs) with floating gate nodes consisting of Au nanoparticles on top of the nanowire channels were investigated in this study. Au nanoparticles, synthesized by a thermal deposition of Au thin film and by a subsequent thermal annealing procedure, were embedded in between Al2O3 tunneling and control gate layers deposited on ZnO nanowire channels. For a representative single ZnO nanowire-based FET with floating gate nodes consisting of Au nanoparticles embedded between Al2O3 layers, its drain current versus gate voltage (I(DS)-V(DS)) characteristics for a double sweep in the gate voltage range from -4 to 4 V exhibit a clockwise hysteresis loop with a threshold voltage shift of deltaV(th) = 1.6 V, resulting from the tunneling of the charge carriers from the ZnO nanowire channel into the Au nanoparticles. In addition, the charge storage characteristic of threshold voltage shift with the elapsed time observed in this FET is also discussed in this paper.

A 60-Hz sinusoidal magnetic field induces apoptosis of prostate cancer cells through reactive oxygen species.

PURPOSE: To explore the effects of power frequency magnetic fields (MF) on cell growth in prostate cancer, DU145, PC3, and LNCaP cells were examined in vitro. MATERIALS AND METHODS: The cells were exposed to various intensities and durations of 60-Hz sinusoidal MF in combination with various serum concentrations in the media. To analyze MF effects on cell growth, cell counting, trypan blue exclusion assay, Western blot analysis, flow cytometry, enzyme-linked immunosorbent assay (ELISA), semi-quantitative reverse transcriptase-polymerase chain reaction (RT-PCR), fluorescence microscopy, and spectrofluorometry were used. RESULTS: MF exposure induced significant cell growth inhibition and apoptosis in an intensity- and time-dependent manner, in which cell cycle arrest, cleaved Caspase-3, and reactive oxygen species (ROS) increased. Pretreatment with a Caspase-3 inhibitor or antioxidant, N-acetyl-L-cysteine (NAC), significantly attenuated MF-induced cell growth inhibition and cell death. Media replacement experiments failed to show any notable change in the MF effects. CONCLUSIONS: These results demonstrate 60-Hz sinusoidal MF-activated cell growth inhibition of prostate cancer in vitro. Apoptosis together with cell cycle arrest were the dominant causes of the MF-elicited cell growth inhibition, mediated by MF-induced ROS. These results suggest that a possibility of using 60-Hz MF in radiation therapy of prostate cancer could usefully be investigated.

NOT and NAND logic circuits composed of top-gate ZnO nanowire field-effect transistors with high-k Al(2)O(3) gate layers.

Electrical characteristics of NOT and NAND logic circuits fabricated using top-gate ZnO nanowire field-effect transistors (FETs) with high-k Al(2)O(3) gate layers were investigated in this study. To form a NOT logic circuit, two identical FETs whose I(on)/I(off) ratios were as high as ∼10(8) were connected in series in a single ZnO nanowire channel, sharing a common source electrode. Its voltage transfer characteristics exhibited an inverting operation and its logic swing was 98%. In addition, the characteristics of a NAND logic circuit composed of three top-gate FETs connected in series in a single nanowire channel are discussed in this paper.

ZnO nanowire-based nano-floating gate memory with Pt nanocrystals embedded in Al(2)O(3) gate oxides.

The memory characteristics of ZnO nanowire-based nano-floating gate memory (NFGM) with Pt nanocrystals acting as the floating gate nodes were investigated in this work. Pt nanocrystals were embedded between Al(2)O(3) tunneling and control oxide layers deposited on ZnO nanowire channels. For a representative ZnO nanowire-based NFGM with embedded Pt nanocrystals, a threshold voltage shift of 3.8 V was observed in its drain current versus gate voltage (I(DS)-V(GS)) measurements for a double sweep of the gate voltage, revealing that the deep effective potential wells built into the nanocrystals provide our NFGM with a large charge storage capacity. Details of the charge storage effect observed in this memory device are discussed in this paper.

Fabrication and device characterization of omega-shaped-gate ZnO nanowire field-effect transistors.

Omega-shaped-gate (OSG) nanowire-based field effect transistors (FETs) have attracted a great deal of attention recently, because theoretical simulations predicted that they should have a higher device performance than nanowire-based FETs with other gate geometries. OSG FETs with channels composed of ZnO nanowires were successfully fabricated in this study using photolithographic processes. In the OSG FETs fabricated on oxidized Si substrates, the channels composed of ZnO nanowires with diameters of about 110 nm are coated with Al(2)O(3) using atomic layer deposition, which surrounds the channels and acts as a gate dielectric. About 80% of the surfaces of the nanowires coated with Al(2)O(3) are covered with the gate metal to form OSG FETs. A representative OSG FET fabricated in this study exhibits a mobility of 30.2 cm(2)/ (V s), a peak transconductance of 0.4 muS (V(g) = -2.2 V), and an I(on)/I(off) ratio of 10(7). To the best of our knowledge, the value of the I(on)/I(off) ratio obtained from this OSG FET is higher than that of any of the previously reported nanowire-based FETs. Its mobility, peak transconductance, and I(on)/I(off) ratio are remarkably enhanced by 3.5, 32, and 10(6) times, respectively, compared with a back-gate FET with the same ZnO nanowire channel as utilized in the OSG FET.