Screening mammography performance according to breast density: a comparison between radiologists versus standalone intelligence detection.

BACKGROUND: Artificial intelligence (AI) algorithms for the independent assessment of screening mammograms have not been well established in a large screening cohort of Asian women. We compared the performance of screening digital mammography considering breast density, between radiologists and AI standalone detection among Korean women. METHODS: We retrospectively included 89,855 Korean women who underwent their initial screening digital mammography from 2009 to 2020. Breast cancer within 12 months of the screening mammography was the reference standard, according to the National Cancer Registry. Lunit software was used to determine the probability of malignancy scores, with a cutoff of 10% for breast cancer detection. The AI's performance was compared with that of the final Breast Imaging Reporting and Data System category, as recorded by breast radiologists. Breast density was classified into four categories (A-D) based on the radiologist and AI-based assessments. The performance metrics (cancer detection rate [CDR], sensitivity, specificity, positive predictive value [PPV], recall rate, and area under the receiver operating characteristic curve [AUC]) were compared across breast density categories. RESULTS: Mean participant age was 43.5 ± 8.7 years; 143 breast cancer cases were identified within 12 months. The CDRs (1.1/1000 examination) and sensitivity values showed no significant differences between radiologist and AI-based results (69.9% [95% confidence interval [CI], 61.7-77.3] vs. 67.1% [95% CI, 58.8-74.8]). However, the AI algorithm showed better specificity (93.0% [95% CI, 92.9-93.2] vs. 77.6% [95% CI, 61.7-77.9]), PPV (1.5% [95% CI, 1.2-1.9] vs. 0.5% [95% CI, 0.4-0.6]), recall rate (7.1% [95% CI, 6.9-7.2] vs. 22.5% [95% CI, 22.2-22.7]), and AUC values (0.8 [95% CI, 0.76-0.84] vs. 0.74 [95% CI, 0.7-0.78]) (all P < 0.05). Radiologist and AI-based results showed the best performance in the non-dense category; the CDR and sensitivity were higher for radiologists in the heterogeneously dense category (P = 0.059). However, the specificity, PPV, and recall rate consistently favored AI-based results across all categories, including the extremely dense category. CONCLUSIONS: AI-based software showed slightly lower sensitivity, although the difference was not statistically significant. However, it outperformed radiologists in recall rate, specificity, PPV, and AUC, with disparities most prominent in extremely dense breast tissue.

Concordant and discordant breast density patterns by different approaches for assessing breast density and breast cancer risk.

PURPOSE: To examine the discrepancy in breast density assessments by radiologists, LIBRA software, and AI algorithm and their association with breast cancer risk. METHODS: Among 74,610 Korean women aged ≥ 34 years, who underwent screening mammography, density estimates obtained from both LIBRA and the AI algorithm were compared to radiologists using BI-RADS density categories (A-D, designating C and D as dense breasts). The breast cancer risks were compared according to concordant or discordant dense breasts identified by radiologists, LIBRA, and AI. Cox-proportional hazards models were used to determine adjusted hazard ratios (aHRs) [95% confidence intervals (CIs)]. RESULTS: During a median follow-up of 9.9 years, 479 breast cancer cases developed. Compared to the reference non-dense breast group, the aHRs (95% CIs) for breast cancer were 2.37 (1.68-3.36) for radiologist-classified dense breasts, 1.30 (1.05-1.62) for LIBRA, and 2.55 (1.84-3.56) for AI. For different combinations of breast density assessment, aHRs (95% CI) for breast cancer were 2.40 (1.69-3.41) for radiologist-dense/LIBRA-non-dense, 11.99 (1.64-87.62) for radiologist-non-dense/LIBRA-dense, and 2.99 (1.99-4.50) for both dense breasts, compared to concordant non-dense breasts. Similar trends were observed with radiologists/AI classification: the aHRs (95% CI) were 1.79 (1.02-3.12) for radiologist-dense/AI-non-dense, 2.43 (1.24-4.78) for radiologist-non-dense/AI-dense, and 3.23 (2.15-4.86) for both dense breasts. CONCLUSION: The risk of breast cancer was highest in concordant dense breasts. Discordant dense breast cases also had a significantly higher risk of breast cancer, especially when identified as dense by either AI or LIBRA, but not radiologists, compared to concordant non-dense breast cases.

Ultrathin Serpentine Insulation Layer Architecture for Ultralow Power Gas Sensor.

Toxic gases have surreptitiously influenced the health and environment of contemporary society with their odorless/colorless characteristics. As a result, a pressing need for reliable and portable gas-sensing devices has continuously increased. However, with their negligence to efficiently microstructure their bulky supportive layer on which the sensing and heating materials are located, previous semiconductor metal-oxide gas sensors have been unable to fully enhance their power efficiency, a critical factor in power-stringent portable devices. Herein, an ultrathin insulation layer with a unique serpentine architecture is proposed for the development of a power-efficient gas sensor, consuming only 2.3 mW with an operating temperature of 300 °C (≈6% of the leading commercial product). Utilizing a mechanically robust serpentine design, this work presents a fully suspended standalone device with a supportive layer thickness of only ≈50 nm. The developed gas sensor shows excellent mechanical durability, operating over 10 000 on/off cycles and ≈2 years of life expectancy under continuous operation. The gas sensor detected carbon monoxide concentrations from 30 to 1 ppm with an average response time of ≈15 s and distinguishable sensitivity to 1 ppm (ΔR/R0 = 5%). The mass-producible fabrication and heating efficiency presented here provide an exemplary platform for diverse power-efficient-related devices.

Ultrathin Crystalline Silicon Nano and Micro Membranes with High Areal Density for Low-Cost Flexible Electronics.

Ultrathin crystalline silicon is widely used as an active material for high-performance, flexible, and stretchable electronics, from simple passive and active components to complex integrated circuits, due to its excellent electrical and mechanical properties. However, in contrast to conventional silicon wafer-based devices, ultrathin crystalline silicon-based electronics require an expensive and rather complicated fabrication process. Although silicon-on-insulator (SOI) wafers are commonly used to obtain a single layer of crystalline silicon, they are costly and difficult to process. Therefore, as an alternative to SOI wafers-based thin layers, here, a simple transfer method is proposed for printing ultrathin multiple crystalline silicon sheets with thicknesses between 300 nm to 13 µm and high areal density (>90%) from a single mother wafer. Theoretically, the silicon nano/micro membrane can be generated until the mother wafer is completely consumed. In addition, the electronic applications of silicon membranes are successfully demonstrated through the fabrication of a flexible solar cell and flexible NMOS transistor arrays.

Gas-Phase Alkali Metal-Assisted MOCVD Growth of 2D Transition Metal Dichalcogenides for Large-Scale Precise Nucleation Control.

Advances in large-area and high-quality 2D transition metal dichalcogenides (TMDCs) growth are essential for semiconductor applications. Here, the gas-phase alkali metal-assisted metal-organic chemical vapor deposition (GAA-MOCVD) of 2D TMDCs is reported. It is determined that sodium propionate (SP) is an ideal gas-phase alkali-metal additive for nucleation control in the MOCVD of 2D TMDCs. The grain size of MoS2 in the GAA-MOCVD process is larger than that in the conventional MOCVD process. This method can be applied to the growth of various TMDCs (MoS2 , MoSe2 , WSe2 , and WSe2 ) and the generation of large-scale continuous films. Furthermore, the growth behaviors of MoS2 under different SP and oxygen injection time conditions are systematically investigated to determine the effects of SP and oxygen on nucleation control in the GAA-MOCVD process. It is found that the combination of SP and oxygen increases the grain size and nucleation suppression of MoS2 . Thus, the GAA-MOCVD with a precise and controllable supply of a gas-phase alkali metal and oxygen allows achievement of optimum growth conditions that maximizes the grain size of MoS2 . It is expected that GAA-MOCVD can provide a way for batch fabrication of large-scale atomically thin electronic devices based on 2D semiconductors.

Switchable, Tunable, and Directable Exciton Funneling in Periodically Wrinkled WS2.

The extreme elastic strain of monolayer transition metal dichalcogenides provides an ideal platform to achieve efficient exciton funneling via local strain modulation; however, studies conducted thus far have focused on the use of substrates with fixed strain profiles. We prepared 1L-WS2 on a flexible substrate such that the formation of topographic wrinkles could be switched on or off, and the depth or the direction of the wrinkle could be modified by external strain, thereby providing full control of the periodic undulation of the band gap profile of 1L-WS2 in the range 0-57 meV. Nanoscale photoluminescence (PL) imaging unambiguously evinced that the photoexcited excitons of 1L-WS2 were accumulated at the top regions of the wrinkles with less band gap than the valley region. Our results of broad tunability of the two-dimensional (2D) exciton funneling suggest a promising route to control exciton drift for enhanced optoelectronic performances and future 2D exciton devices.

Arrayed MoS2 -In0.53 Ga0.47 As van der Waals Heterostructure for High-Speed and Broadband Detection from Visible to Shortwave-Infrared Light.

A high-speed and broadband 5 × 5 photodetector array based on MoS2 /In0.53 Ga0.47 As heterojunction is successfully demonstrated to take full advantage of the type-II band-aligned multilayer MoS2 /In0.53 Ga0.47 As. The fabricated devices exhibit good uniformity in the Raman spectrum and clear rectifying characteristics. The fabricated MoS2 /In0.53 Ga0.47 As photodetectors show good optical performances at a broad wavelength range showing high responsivities corresponding to the detectivity of ≈1010 Jones at -3 V for the incident broadband light from 400 to 1550 nm. A very fast photoresponse is also obtained with a small rise/fall time in the order of microseconds both for visible (638 nm) and shortwave infrared (1310 nm). Finally, the image scanning properties of MoS2 /In0.53 Ga0.47 As devices are demonstrated for visible and infrared light, indicating that the suggested device is one of the promising options for future broadband imager, which can be integrated on the focal plane arrays (FPAs).

Mechanoluminescent, Air-Dielectric MoS2 Transistors as Active-Matrix Pressure Sensors for Wide Detection Ranges from Footsteps to Cellular Motions.

Tactile pressure sensors as flexible bioelectronic devices have been regarded as the key component for recently emerging applications in electronic skins, health-monitoring devices, or human-machine interfaces. However, their narrow range of sensible pressure and their difficulty in forming high integrations represent major limitations for various potential applications. Herein, we report fully integrated, active-matrix arrays of pressure-sensitive MoS2 transistors with mechanoluminescent layers and air dielectrics for wide detectable range from footsteps to cellular motions. The inclusion of mechanoluminescent materials as well as air spaces can increase the sensitivity significantly over entire pressure regimes. In addition, the high integration capability of these active-matrix sensory circuitries can enhance their spatial resolution to the level sufficient to analyze the pressure distribution in a single cardiomyocyte. We envision that these wide-range pressure sensors will provide a new strategy toward next-generation electronics at biomachine interfaces to monitor various mechanical and biological phenomena at single-cell resolution.

Inhibiting casein kinase 2 overcomes paclitaxel resistance in gastric cancer.

PURPOSE: Casein kinase (CK) 2 activation has been implicated in the proliferation of various tumor types and resistance to chemotherapy. We investigated the mechanistic basis for the association between CK2 activation and paclitaxel resistance in a gastric cancer (GC). EXPERIMENTAL DESIGN: CK2 expression was evaluated in 59 advanced GC patients treated with paclitaxel as the second-line therapy. The efficacy of a CK2 inhibitor, CX-4945, and paclitaxel was evaluated in GC cell lines and a xenograft model. RESULTS: Patients with high CK2 expression (29/59, 39%) showed lower disease control rates (47.7% vs. 72.3%, p = 0.017) and shorter progression-free survival (2.8 vs. 4.8 months, p = 0.009) than patients with low CK2 expression. CK2 protein expression was associated with sensitivity to paclitaxel in 49 GC cell lines. Combination therapy with CX-4945 and paclitaxel exerted synergistic antiproliferative effects and inhibited the downregulation of phosphatidylinositol 3-kinase/AKT signaling in SNU-1 cells. In the SNU-1 xenograft model, the combination treatment was significantly superior to either single agent, suppressing tumor growth without notable toxicities. CONCLUSIONS: These results demonstrated that CK2 activation was related to paclitaxel resistance and that CX-4945 in combination with paclitaxel could be used as a potential treatment for paclitaxel resistance in GC.

Forty-nine gastric cancer cell lines with integrative genomic profiling for development of c-MET inhibitor.

Receptor tyrosine kinase MET (c-MET) has received considerable attention as a potential target for gastric cancer (GC) therapy and a number of c-MET inhibitors have been developed. For successful drug development, proper preclinical studies especially using patient derived cancer cell lines are very important. We profiled MET and MET-related characteristics in 49 GC cell lines to utilize them as models in preclinical studies of GC. Forty-nine cell lines were analyzed for genetic, biological, and molecular status to characterize MET and MET-related molecules. Four c-MET inhibitors were tested to elucidate the dependency on MET pathway in the 49 GC cell lines. Six of 49 cell lines were MET amplified with overexpression of c-MET and p-MET. The variants of MET were not associated with c-MET expression or amplification. Hs746T showed an exon 14 deletion in conjunction with MET amplification. The cell lines were divided into 6 MET amplified, 2 c-MET overexpressed, 2 hepatocyte growth factor (HGF) overexpressed, and 39 MET-negative subgroups. Except tivantinib, the c-MET inhibitors showed higher inhibition (%) in MET amplified than in MET nonamplified cell lines that MET amplified cell lines showed MET pathway dependency. However, the c-MET overexpressed and HGF overexpressed cell lines showed moderate dependency on MET pathway. Well-characterized cell lines are very important in studying drug development. Our 49 GC cell lines had various characteristics of MET and MET-related molecules and MET pathway dependency. These provide a promising platform for development of various RTK inhibitors including c-MET inhibitors.

Light-soaking effects and capacitance profiling in Cu(In,Ga)Se2 thin-film solar cells with chemical-bath-deposited ZnS buffer layers.

We fabricated Cu(In,Ga)Se2 (CIGS) solar cells with chemical-bath deposited (CBD) ZnS buffer layers with different deposition times. The conversion efficiency and the fill factor of the CIGS solar cells reveal a strong dependence on the deposition time of CBD-ZnS films. In order to understand the detailed relationship between the heterojunction structure and the electronic properties of CIGS solar cells with different deposition times of CBD-ZnS films, capacitance-voltage (C-V) profiling measurements with additional laser illumination were performed. The light-soaking effects on CIGS solar cells with a CBD-ZnS buffer layer were investigated in detail using current density-voltage (J-V) and C-V measurements with several different lasers with different emission wavelengths. After light-soaking, the conversion efficiency changed significantly and the double diode feature in J-V curves disappeared. We explain that the major reason for the improvement of efficiency by light-soaking is due to the fact that negatively charged and highly defective vacancies in the CIGS absorber near the interface of CBD-ZnS/CIGS were formed and became neutral due to carriers generated by ultra-violet absorption in the buffer layer.

PINCH-2 presents functional copy number variation and suppresses migration of colon cancer cells by paracrine activity.

In recent years, characterization of cancer and its environment has become necessary. However, studies of the cancer microenvironment remain insufficient. Copy number variations (CNVs) occur in 40% of cancer-related genes, but few studies have reported the correlation between CNVs in morphologically normal tissues adjacent to cancer and cancer progression. In this study, we evaluated cancer cell migration and invasion according to the genetic differences between cancer tissues and their surrounding normal tissues. To study the field cancerization effect, we screened 89 systemic metastasis-related CNVs from morphologically normal tissues adjacent to colon cancers. Among these CNVs, LIM and senescent cell antigen-like domain 2 (PINCH-2) showed copy number amplification and upregulation of mRNA in the nonrelapsed group compared to the systemic relapse group. PINCH-2 expression in colon cancer cells was lower than that in normal epithelial colon cells at both the protein and mRNA levels. Suppression of PINCH-2 resulted in decreased formation of the PINCH-2-IPP (PINCH-2, integrin-linked kinase and α-parvin) complex and reciprocally increased formation of the PINCH-1-IPP complex. Although PINCH-2 expression of survival pathway-related proteins (Akt and phospho-Akt) did not change upon suppression of PINCH-2 expression, cell migration-related proteins [matrix-metalloproteinase (MMP)-9 and -11] were upregulated through autocrine and paracrine activation. Thus, PINCH-2 participates in decreased systemic recurrence by competitively regulating IPP complex formation with PINCH-1, thereby suppressing autocrine and paracrine effects on motility in colon cancer. This genetic change in morphologically normal tissue suggests a field cancerization effect of the tumor microenvironment in cancer progression.

High KLF4 level in normal tissue predicts poor survival in colorectal cancer patients.

BACKGROUND: Krüppel-like factor 4 (KLF4) is involved in many important cellular processes such as growth, development, differentiation, proliferation, and apoptosis. The purpose of this study was to determine the significance of KLF4 in both tumors and normal tissues of patients with colorectal cancer (CRC). METHODS: Between January 2003 and June 2005, 125 patients with CRC receiving treatment at the Yonsei Cancer Center were selected. We examined the mRNA level of the KLF4 gene in primary CRC specimens and matched normal colon tissues using real-time RT-PCR. Correlation of survival with clinicopathological parameters, including KLF4 level, was investigated with univariate and multivariate analyses. RESULTS: CRC tissue had a significantly lower KLF4 level when compared with matched normal tissues (KLF4 in tumors: 2007 ± 1531 copies/µl, KLF4 in normal tissues: 6586 ± 2834 copies/µl; P < 0.0001). However, there was a correlation between the KLF4 level in tumors and normal tissues. Patients with a high KLF4 level in matched normal tissues were more likely than those with a low KLF4 level to develop recurrence and had poorer overall survival (P = 0.005). Therefore, the KLF4 level in the normal tissue of individuals was associated with prognosis of individuals. CONCLUSIONS: Our data suggest that KLF4 mRNA expression level in normal tissues and tumors may be a useful prognostic marker in patients with CRC.

An experimental and numerical study on adhesion force at the nanoscale.

Adhesion has attracted great interest in science and engineering especially in the field pertaining to nano-science because every form of physical contact is fundamentally a macroscopic observation of interactions between nano-asperities under the adhesion phenomenon. Despite its importance, no practical adhesion prediction model has been developed due to the complexity of examining contact between nano-asperities. Here, we scrutinized the contact phenomenon and developed a contact model, reflecting the physical sequence in which adhesion develops. For the first time ever, our model analyzes the adhesion force and contact properties, such as separation distance, contact location, actual contact area, and the physical deformation of the asperities, between rough surfaces. Through experiments using atomic force microscopy, we demonstrated a low absolute percentage error of 2.8% and 6.55% between the experimental and derived data for Si-Si and Mo-Mo contacts, respectively, and proved the accuracy and practicality of our model in the analysis of the adhesion phenomenon.

Fully bioresorbable hybrid opto-electronic neural implant system for simultaneous electrophysiological recording and optogenetic stimulation.

Bioresorbable neural implants based on emerging classes of biodegradable materials offer a promising solution to the challenges of secondary surgeries for removal of implanted devices required for existing neural implants. In this study, we introduce a fully bioresorbable flexible hybrid opto-electronic system for simultaneous electrophysiological recording and optogenetic stimulation. The flexible and soft device, composed of biodegradable materials, has a direct optical and electrical interface with the curved cerebral cortex surface while exhibiting excellent biocompatibility. Optimized to minimize light transmission losses and photoelectric artifact interference, the device was chronically implanted in the brain of transgenic mice and performed to photo-stimulate the somatosensory area while recording local field potentials. Thus, the presented hybrid neural implant system, comprising biodegradable materials, promises to provide monitoring and therapy modalities for versatile applications in biomedicine.

Area-selective atomic layer deposition on 2D monolayer lateral superlattices.

The advanced patterning process is the basis of integration technology to realize the development of next-generation high-speed, low-power consumption devices. Recently, area-selective atomic layer deposition (AS-ALD), which allows the direct deposition of target materials on the desired area using a deposition barrier, has emerged as an alternative patterning process. However, the AS-ALD process remains challenging to use for the improvement of patterning resolution and selectivity. In this study, we report a superlattice-based AS-ALD (SAS-ALD) process using a two-dimensional (2D) MoS2-MoSe2 lateral superlattice as a pre-defining template. We achieved a minimum half pitch size of a sub-10 nm scale for the resulting AS-ALD on the 2D superlattice template by controlling the duration time of chemical vapor deposition (CVD) precursors. SAS-ALD introduces a mechanism that enables selectivity through the adsorption and diffusion processes of ALD precursors, distinctly different from conventional AS-ALD method. This technique facilitates selective deposition even on small pattern sizes and is compatible with the use of highly reactive precursors like trimethyl aluminum. Moreover, it allows for the selective deposition of a variety of materials, including Al2O3, HfO2, Ru, Te, and Sb2Se3.

Boltzmann Switching MoS2 Metal-Semiconductor Field-Effect Transistors Enabled by Monolithic-Oxide-Gapped Metal Gates at the Schottky-Mott Limit.

A gate stack that facilitates a high-quality interface and tight electrostatic control is crucial for realizing high-performance and low-power field-effect transistors (FETs). However, when constructing conventional metal-oxide-semiconductor structures with two-dimensional (2D) transition metal dichalcogenide channels, achieving these requirements becomes challenging due to inherent difficulties in obtaining high-quality gate dielectrics through native oxidation or film deposition. Here, a gate-dielectric-less device architecture of van der Waals Schottky gated metal-semiconductor FETs (vdW-SG MESFETs) using a molybdenum disulfide (MoS2) channel and surface-oxidized metal gates such as nickel and copper is reported. Benefiting from the strong SG coupling, these MESFETs operate at remarkably low gate voltages, <0.5 V. Notably, they also exhibit Boltzmann-limited switching behavior featured by a subthreshold swing of ≈60 mV dec-1 and negligible hysteresis. These ideal FET characteristics are attributed to the formation of a Fermi-level (EF) pinning-free gate stack at the Schottky-Mott limit. Furthermore, authors experimentally and theoretically confirm that EF depinning can be achieved by suppressing both metal-induced and disorder-induced gap states at the interface between the monolithic-oxide-gapped metal gate and the MoS2 channel. This work paves a new route for designing high-performance and energy-efficient 2D electronics.

Layer-Controlled Growth of Single-Crystalline 2D Bi2O2Se Film Driven by Interfacial Reconstruction.

As semiconductor scaling continues to reach sub-nanometer levels, two-dimensional (2D) semiconductors are emerging as a promising candidate for the post-silicon material. Among these alternatives, Bi2O2Se has risen as an exceptionally promising 2D semiconductor thanks to its excellent electrical properties, attributed to its appropriate bandgap and small effective mass. However, unlike other 2D materials, growth of large-scale Bi2O2Se films with precise layer control is still challenging due to its large surface energy caused by relatively strong interlayer electrostatic interactions. Here, we present the successful growth of a wafer-scale (∼3 cm) Bi2O2Se film with precise thickness control down to the monolayer level on TiO2-terminated SrTiO3 using metal-organic chemical vapor deposition (MOCVD). Scanning transmission electron microscopy (STEM) analysis confirmed the formation of a [BiTiO4]1- interfacial structure, and density functional theory (DFT) calculations revealed that the formation of [BiTiO4]1- significantly reduced the interfacial energy between Bi2O2Se and SrTiO3, thereby promoting 2D growth. Additionally, spectral responsivity measurements of two-terminal devices confirmed a bandgap increase of up to 1.9 eV in monolayer Bi2O2Se, which is consistent with our DFT calculations. Finally, we demonstrated high-performance Bi2O2Se field-effect transistor (FET) arrays, exhibiting an excellent average electron mobility of 56.29 cm2/(V·s). This process is anticipated to enable wafer-scale applications of 2D Bi2O2Se and facilitate exploration of intriguing physical phenomena in confined 2D systems.

Sub-10 fJ/bit radiation-hard nanoelectromechanical non-volatile memory.

With the exponential growth of the semiconductor industry, radiation-hardness has become an indispensable property of memory devices. However, implementation of radiation-hardened semiconductor memory devices inevitably requires various radiation-hardening technologies from the layout level to the system level, and such technologies incur a significant energy overhead. Thus, there is a growing demand for emerging memory devices that are energy-efficient and intrinsically radiation-hard. Here, we report a nanoelectromechanical non-volatile memory (NEM-NVM) with an ultra-low energy consumption and radiation-hardness. To achieve an ultra-low operating energy of less than 10 [Formula: see text], we introduce an out-of-plane electrode configuration and electrothermal erase operation. These approaches enable the NEM-NVM to be programmed with an ultra-low energy of 2.83 [Formula: see text]. Furthermore, due to its mechanically operating mechanisms and radiation-robust structural material, the NEM-NVM retains its superb characteristics without radiation-induced degradation such as increased leakage current, threshold voltage shift, and unintended bit-flip even after 1 Mrad irradiation.