Abstract:
The present disclosure relates to a method and apparatus for automatic thoracic organ segmentation. The method includes: receiving three-dimensional (3D) images obtained by a computed tomography (CT) system; processing the 3D images to have the same spatial resolution and matrix size; building a two-stage deep learning framework using convolutional neural networks (CNNs) for organ segmentation; adapting the deep learning framework to be compatible with incomplete training data; improving the CNNs upon arrival of new training data; post-processing the output from the deep learning framework to obtain final organ segmentation.
Abstract:
A method for packaging an integrated circuit (IC) device in which conventional manufacturing steps of mechanically bonding a die to a corresponding interconnecting substrate, wire bonding the die, and encapsulating the die in a protective shell are replaced by a single manufacturing step that includes thermally treating an appropriate assembly of parts to both form proper electrical connections for the die in the resulting IC package and cause the molding compound(s) to encapsulate the die in a protective enclosure.
Abstract:
A method of providing graphics data, comprising generating a first set of data vectors specifying geometrical characteristics of a graphical object in a first digital picture, generating a second set of data vectors specifying geometrical characteristics of the graphical object in a second digital picture to be displayed after the first digital picture, generating a parameter set comprising information specifying intermediate geometrical characteristics of the graphical object based on the geometrical characteristics of the graphical object in the first digital picture, and of the graphical object in the second digital picture, wherein the intermediate geometrical characteristics are geometrical characteristics of the graphical object in at least one third digital picture to be displayed after the first digital picture and before the second digital picture, and generating at least one data file comprising the first set of data vectors, the second set of data vectors and the parameter set.
Abstract:
An apparatus for molding a semiconductor device includes an upper mold chase and a lower mold chase. The mold chases are capable of being aligned with each other, forming spaced cavities for receiving a lead frame array that includes semiconductor dies for encapsulation. The cavities are aligned in spaced, vertical columns and gates are provided at the opening of each column of cavities. A molding compound is passed through the gates and flows uninterrupted through each cavity and encapsulates the semiconductor dies.
Abstract:
A system and method of optimizing delivery of a radiation therapy treatment. The system optimizes treatment delivery in real-time to take into account a variety of factors, such as patient anatomical and physiological changes (e.g., respiration and other movement, etc.), and machine configuration changes (e.g., beam output factors, couch error, leaf error, etc.).
Abstract:
The present invention discloses a wireless communication method of supporting rapid changes in network topology. The present method may include broadcasting the first protocol message using the first broadcast cycle, the first protocol message being used for establishing a routing information between a node and destination node; and broadcasting the second protocol message using the second broadcast cycle, the second protocol message being used for establishing a link information between the node and a neighbor node, wherein the first broadcast cycle is different from the second broadcast cycle, and the first broadcast cycle is dynamically adjusted depending on a corresponding a detailed message contained in a plurality of segments in the first protocol message so as to reduce a number of broadcast of the first protocol message.
Abstract:
A method of calculating a dose distribution for a patient for use in a radiation therapy treatment plan. The method includes acquiring an image of a volume within the patient, defining a radiation source, and defining a reference plane oriented between the radiation source and the patient. The method also includes generating a radiation therapy treatment plan, wherein the plan includes a plurality of rays that extend between the radiation source and the patient volume, and calculating a three-dimensional dose volume for the patient volume from the plurality of rays that intersect the reference plane without first having to independently calculate a dose distribution on each of the plurality of rays. The method can also include displaying the three-dimensional dose volume.
Abstract:
A system and method of detecting a breathing phase of a patient receiving radiation therapy is disclosed. The method, in one implementation, includes the acts of obtaining a plurality of patient images representing phases of a breathing cycle, delivering radiation to the patient, collecting transmission data of the patient during the delivering radiation, and comparing the transmission data to the plurality of patient images.
Abstract:
The present disclosure relates to a method and apparatus for automatic head and neck organ segmentation. The method includes: prepare a training dataset, wherein the step of preparing a training dataset comprises: receive 3D images covering head and neck region obtained by a CT system or an MRI system; receive metadata for each received 3D images, comprising patient orientation, pixel spacing, slice thickness and matrix size; receive corresponding 3D segmentation labels map for the received 3D images; and process the received 3D images and the corresponding 3D segmentation labels map by transforming to patient coordinate system and resampling to have a fixed spatial resolution and a fixed matrix size; build a deep learning framework using CNN models for organ segmentation; train CNN models using the training dataset and performing data emulation step during training by mirroring the processed 3D images and their corresponding processed 3D segmentation labels map from the training dataset; prepare a testing dataset, wherein the step of preparing a testing dataset comprises: receive 3D images covering head and neck region obtained by a CT system or an MRI system; receive metadata for each received 3D images, comprising patient orientation, pixel spacing, slice thickness and matrix size; and process the received 3D images by transforming to patient coordinate system and resampling to have a fixed spatial resolution and a fixed matrix size; deploy the trained CNN models on testing dataset, wherein the testing step comprises: mirror the processed 3D images in the left-right direction; predict on the processed 3D images and mirrored 3D images with individual prediction outputs as 3D probabilities map for organ segmentation; and improve the segmentation performance by averaging the 3D probabilities map outputs from the processed 3D images and mirrored 3D images; and post-process the 3D probabilities map outputs from the deep learning framework to obtain final 3D segmentation labels map for organ segmentation.
Abstract:
A method for packaging an integrated circuit (IC) device in which conventional manufacturing steps of mechanically bonding a die to a corresponding interconnecting substrate, wire bonding the die, and encapsulating the die in a protective shell are replaced by a single manufacturing step that includes thermally treating an appropriate assembly of parts to both form proper electrical connections for the die in the resulting IC package and cause the molding compound(s) to encapsulate the die in a protective enclosure.