Abstract: An example computer-implemented method for self-supervised training of an image processing model for histopathology images is provided. The example method includes obtaining a reference histopathology image; generating an augmented histopathology image, wherein generating the augmented histopathology image comprises performing, for an input image, at least one of the following augmentations: applying a blur to the input image and injecting noise artifacts into the blurred input image; or cropping a plurality of portions from the input image, wherein the plurality of portions are determined based on a minimum overlap criterion that has been updated over one or more iterations; and training the image processing model based on a similarity of latent representations generated by the image processing model respectively for the reference histopathology image and the augmented histopathology image.

Abstract: A semiconductor structure and method of manufacturing a semiconductor structure are provided. The semiconductor structure includes a substrate and at least one contact plug. The substrate has an epi-layer. The contact plug is formed on the epi-layer and includes a silicide cap disposed on the epi-layer; a conductive pillar disposed on the silicide cap such that the conductive pillar electrically connects to the epi-layer via the silicide cap; and a hybrid liner. The hybrid liner surrounds the conductive pillar and includes a lower portion abutting the silicide cap and having a nitride material and an upper portion abutting the conductive pillar and having an oxidized nitride material. Due to the hybrid liner, a semiconductor structure with increased capacitance and decreased resistivity can be obtained.

Abstract: A head wearable display system comprising a target object detection module receiving multiple image pixels of a first portion and a second portion of a target object, and the corresponding depths; a first light emitter emitting multiple first-eye light signals to display a first-eye virtual image of the first portion and the second portion of the target object for a viewer; a first light direction modifier for respectively varying a light direction of each of the multiple first-eye light signals emitted from the first light emitter; a first collimator; a first combiner, for redirecting and converging the multiple first-eye light signals towards a first eye of the viewer. The first-eye virtual image of the first portion of the target object in a first field of view has a greater number of the multiple first-eye light signals per degree than that of the first-eye virtual image of the second portion of the target object in a second field of view.

Abstract: A machine learning training device is disclosed. The machine learning training device includes a virtual hard anchor generation circuit, a classification circuit and a training circuit. The virtual hard anchor generation circuit is configured to generate several virtual hard anchors according to several easy samples classified into several types. The virtual hard anchors respectively correspond to one of the several types. The classification circuit is configured to classify several hard samples into several types according to virtual hard anchors. Parts of the hard samples classified into several types are several clean hard samples. Another parts of the hard samples that are not classified into several types are several noisy hard samples. The training circuit is configured to perform machine learning training according to several easy samples and several clean hard samples.

Abstract: An automatic processing device for liquid samples includes a sample region, a control module, an image identification device and a centrifuge. The sample region is configured to accommodate a plurality of centrifuge tubes. The control module includes a mechanical module. The mechanical module is configured to unscrew or tighten upper caps of the centrifuge tubes, and is configured to draw liquid from the centrifuge tubes or discharge liquid to the centrifuge tubes. The image identification device is coupled to the control module. The centrifuge is coupled to the control module. The centrifuge is configured to accommodate the centrifuge tubes and perform centrifugal treatment.

Abstract: A hypochlorous acid preparation system is provided. The hypochlorous acid preparation system includes: a hypochlorous acid preparation apparatus comprising: a first inlet, wherein sulfuric acid collected from a clean room located in a semiconductor fabrication plant enters the hypochlorous acid preparation apparatus through the first inlet; a second inlet, wherein sodium hypochlorite solution enters the hypochlorous acid preparation apparatus through the second inlet; a third inlet, wherein deionized water enters the hypochlorous acid preparation apparatus through the third inlet; and an outlet, wherein hypochlorous acid is produced in situ by mixing the sulfuric acid, the sodium hypochlorite solution, and the deionized water and exits the hypochlorous acid preparation apparatus through the outlet.

Abstract: The present invention disclose a medical image-based system for predicting lesion classification and a method thereof. The system comprises a feature data extracting module for providing a raw feature data based on a medical image, and a predicting module for outputting a predicted class and a risk index according to the raw feature data. The predicting module comprises a classification unit for generating the predicted class and a prediction score corresponding thereto according to the raw feature data, and a risk evaluation unit for generating the risk index according to the prediction score. The system provides medical personnels a reference score and a risk index to determine progression of a certain disease.

Abstract: The present invention relates to a method for measuring muscle mass, including: a first selection step, wherein a frame selection information is obtained by using a frame to select a fascia region from a provided computed tomography image under the condition that the window width ranges from 300 HU to 500 HU and the window level ranges from 40 HU to 50 HU, wherein the selected range of the fascia region includes a muscle; and a second selection step, wherein a muscle information of the muscle is obtained by calculating a pixel value in the frame-selected fascia region under the condition that the HU value of the CT image ranges from ?29 HU to 150 HU.

Abstract: An electrode for a lithium-ion battery is provided, which comprises: a current collector; and an electrode material layer disposed on the current collector, wherein the electrode material layer comprises an anode material and a binder, the binder is pectin, its derivative or a combination thereof, and the anode material is selected from the group consisting of lithium vanadium oxide, lithium titanium oxide, lithium iron oxide, graphite, and a combination thereof. In addition, a lithium-ion battery comprising the aforesaid electrode is also provided.

Abstract: An antenna structure and an electronic apparatus are provided. The antenna structure includes a substrate, a first radiation part, and a second radiation part. The substrate has a first surface and a second surface opposite to each other. The first radiation part is disposed on the first surface. The first radiation part is an absorber material. The second radiation part is disposed on the second surface. The second radiation part is coupled to a feeding part. There is a distance between the second radiation part and the first radiation part, so as to excite a first resonance mode through the coupling of the second radiation part to the first radiation part. Accordingly, the specific absorption rate (SAR) value of the electromagnetic wave is reduced.

Abstract: One example method for biomarker detection in digitized pathology samples includes receiving a plurality of image patches corresponding to an image of a pathology slide having a hematoxylin and eosin-stained (“H&E”) stained sample of tissue, each image patch representing a different portion of the image; for each image patch, determining, using a first trained machine learning (“ML”) model, a patch biomarker status; and determining, using a second trained ML model, a tissue sample biomarker status for the sample of tissue based on the patch biomarker statuses of the image patches.

Abstract: A spot type atmospheric pressure plasma device includes a metal casing, a metal electrode, a dielectric layer, and a gas channel. The metal electrode is disposed in an inner space of the metal casing. The dielectric layer is disposed in the inner space and surrounds an outer side surface of the metal electrode. A central area of a bottom of the dielectric layer has a plasma jet, and a bottom of the metal electrode is adjacent to the plasma jet. The gas channel includes a first section, a second section, and a third section. The first section passes through the metal casing and the dielectric layer. The second section is connected to the first section and extends between the dielectric layer and the outer side surface. The third section is connected to the second section, and is configured to direct a working gas to the plasma jet.

Abstract: A wide area atmospheric pressure plasma device includes a metal casing, a metal electrode, and a dielectric layer. The metal casing includes a chamber, at least one gas channel, and a plasma jet channel, in which the plasma jet channel is located under the chamber. The metal electrode is disposed within the chamber, is adjacent to the plasma jet channel, and extends along a length direction of the plasma jet channel. An outlet of the gas channel is adjacent to a bottom of the metal electrode, such that a working gas in the gas channel is sprayed towards the bottom of the metal electrode. The dielectric layer wraps the metal electrode.

Abstract: An antenna structure and an electronic apparatus are provided. The antenna structure includes a substrate, a first radiation part, and a second radiation part. The substrate has a first surface and a second surface opposite to each other. The first radiation part is disposed on the first surface. The first radiation part is an absorber material. The second radiation part is disposed on the second surface. The second radiation part is coupled to a feeding part. There is a distance between the second radiation part and the first radiation part, so as to excite a first resonance mode through the coupling of the second radiation part to the first radiation part. Accordingly, the specific absorption rate (SAR) value of the electromagnetic wave is reduced.

Abstract: A memory cell includes a substrate. A first STI and a second STI are embedded within the substrate. The first STI and the second STI extend along a first direction. An active region is disposed on the substrate and between the first STI and the second STI. A control gate is disposed on the substrate and extends along a second direction. The first direction is different from the second direction. A tunneling region is disposed in the active region overlapping the active region. A first trench is embedded within the tunneling region. Two second trenches are respectively embedded within the first STI and the second STI. The control gate fills in the first trench and the second trenches. An electron trapping stack is disposed between the tunneling region and the control gate.

Abstract: A memory cell includes a substrate. A first STI and a second STI are embedded within the substrate. The first STI and the second STI extend along a first direction. An active region is disposed on the substrate and between the first STI and the second STI. A control gate is disposed on the substrate and extends along a second direction. The first direction is different from the second direction. A tunneling region is disposed in the active region overlapping the active region. A first trench is embedded within the tunneling region. Two second trenches are respectively embedded within the first STI and the second STI. The control gate fills in the first trench and the second trenches. An electron trapping stack is disposed between the tunneling region and the control gate.

Abstract: A memory cell includes a substrate. A first STI and a second STI are embedded within the substrate. The first STI and the second STI extend along a first direction. An active region is disposed on the substrate and between the first STI and the second STI. A control gate is disposed on the substrate and extends along a second direction. The first direction is different from the second direction. A tunneling region is disposed in the active region overlapping the active region. A first trench is embedded within the tunneling region. Two second trenches are respectively embedded within the first STI and the second STI. The control gate fills in the first trench and the second trenches. An electron trapping stack is disposed between the tunneling region and the control gate.

Abstract: A memory cell includes a substrate. A first STI and a second STI are embedded within the substrate. The first STI and the second STI extend along a first direction. An active region is disposed on the substrate and between the first STI and the second STI. A control gate is disposed on the substrate and extends along a second direction. The first direction is different from the second direction. A tunneling region is disposed in the active region overlapping the active region. A first trench is embedded within the tunneling region. Two second trenches are respectively embedded within the first STI and the second STI. The control gate fills in the first trench and the second trenches. An electron trapping stack is disposed between the tunneling region and the control gate.

Abstract: A method for fabricating gate structures includes providing a substrate, configured to have a first region and a second region. Dummy gate structures are formed on the substrate at the first and second regions, wherein each of the dummy gate structures has a first gate insulating layer on the substrate and a dummy gate on the first gate insulating layer. An inter-layer dielectric layer is formed over the dummy gate structures. The inter-layer dielectric layer is polished to expose all of the dummy gates. The dummy gates are removed. The first gate insulating layer at the second region is removed. A second gate insulating layer is formed on the substrate at the second region, wherein the first gate insulating layer is thicker than the second insulating layer. Metal gates are formed on the first and the second insulating layer.

Abstract: A non-volatile memory device is provided. The non-volatile memory device includes a substrate, a first gate structure disposed on the substrate, a second gate structure disposed on the substrate, and a memory gate structure disposed on the substrate and between the first gate structure and the second gate structure. The memory gate structure at least covers the first gate structure and the second gate structure. The memory gate structure includes a charge storage layer disposed on the substrate and a memory gate layer disposed on the charge storage layer.