HKW Intellectual Property PartG mbB

Pulse field ablation (PFA) guidance is generated from intracardiac echocardiogram (ICE) imaging. The relative position of the PFA device (450) and the tissue is determined. The treatment placement with respect to the tissue for repetitive application is detected. By finding the relative position and/or treatment placement over time, ICE imaging in combination with image processing may provide visual guidance rather than just imaging or just catheter tracking for PFA. The resulting PFA procedure may be more likely complete and/or optimized.

The invention relates to a computer-implemented method for training an artificial intelligence model for segmenting blood vessels in an angiography image comprising preprocessing a training data set, the training data set comprising angiography images depicting an arterial phase and/or a venous phase of the blood vessels, to decrease a noise level and/or to enhance a contrast in the angiography images, providing the angiography images depicting an arterial phase to a first artificial intelligence model, providing the angiography images depicting a venous phase to a second artificial intelligence model, training the first artificial intelligence model for segmenting arteries in the angiography images, training the second artificial intelligence model for segmenting veins in the angiography images.

The present application provides a medical image processing method and system, a device and a medium, for acquiring raw data; subjecting the raw data to image reconstruction, to generate an image to be processed; preprocessing the image to be processed; inputting the preprocessed image to be processed into a trained image optimization model to obtain a target image, such that a resolution of the target image is higher than a resolution of the image to be processed; wherein the image optimization model is obtained by training a diffusion model, a training process of the diffusion model having high-dose high-resolution image data as a starting point; the raw data originates in a first device, the high-dose high-resolution image data used to train the diffusion model originates in a second device, and the second device has better image processing capability than the first device. The present application can significantly improve medical image quality at a low cost.

Embodiments of the present invention disclose a tube voltage adjustment method and apparatus, an X-ray system, and a computer-readable storage medium. The method comprises: at the end of a present exposure, obtaining an X-ray image acquired by the present exposure, and obtaining an average grayscale of the X-ray image; obtaining a first ratio between the average grayscale and a target grayscale of this X-ray examination; obtaining an adjustment direction and an adjustment extent of tube voltage of an X-ray tube according to the first ratio; and adjusting a present tube voltage according to the adjustment direction and adjustment extent of tube voltage of the X-ray tube. Embodiments of the present invention increase the speed of X-ray examination.

A draping system for a robotic drive having an elongated drive body and an articulating support arm supporting the elongated drive body, the draping system comprising: a first tubular drape configured to cover a first portion of the elongated drive body; a second tubular drape configured to cover a second portion of the elongated drive body, the second tubular drape being separate from the first tubular drape; and a third drape configured to cover the articulating support arm, the third drape being separate from the first tubular drape and the second tubular drape.

A framework for enhancing isosurfaces using virtual light sources in medical images. The framework positions (204) one or more virtual light sources within the structure of interest in medical image data. Illumination effects (206) generated by the one or more virtual light sources are determined. Isosurface rendering of the medical image data using the illumination effects is then performed (208) to generate a rendered image for display.

A framework for assisted placement of virtual light sources in medical image data. The framework receives (204) medical image data of at least one structure of interest and calculates (206) at least one geometrical feature of the at least one structure of interest based on the medical image data. An optimal position of at least one virtual light source is determined (208) near or within the structure of interest using the geometrical feature. Volume rendering (210) of the medical image data is performed using the at least one virtual light source at the optimal position to generate a projection image.

The present invention relates to an implantable medical device, e.g. a stent, particularly adapted for restenosis identification. Furthermore, the present invention relates to a system, particularly adapted for restenosis identification, the system including such an implantable medical device and a number of readers being coupeable with the implantable medical device. An implantable medical device comprises a structure having a cylindrical shape and an inner lumen, and a number of radio transponders, the respective radio transponder being configured to transmit a radio signal having a certain frequency to a receiver, particularly adapted for restenosis identification. Using the radio transponders, it is advantageously possible to monitor the health of the implantable medical device.

The invention relates to a computer-implemented method for training an artificial intelligence model for segmenting a thalamic nucleus in a radiology image, comprising obtaining a plurality of radiology images comprising a representation of thalamic nuclei, preprocessing the obtained plurality of radiology images, wherein the preprocessing comprises generating a plurality of track density images, TDI, based on the generated plurality of radiology images and training the artificial intelligence model based on at least a first portion of the generated plurality of TDIs.

The present invention generally relates to the field of data processing systems, and more particularly relates to methods and systems for generating domain-knowledge enriched engineering diagrams corresponding to assets in a technical installation. The method includes receiving engineering diagrams corresponding to one or more assets (104) installed in one or more technical installations (120), from data sources, and determining schematic elements and topologies using visual recognition. Further, the method includes identifying correlations, associations, classifications, and patterns based on contextual features. A structured large language model (LLM) prompt is created to validate the relationships, generating an LLM response. The method further includes enhancing schematic features using semantic rules, ontologies, and hierarchical taxonomies, to generate enriched engineering diagrams for display on electronic devices (108).