Philips Intellectual Property & Standards

A system includes a processor configured for communication with an ultrasound probe. The processor is configured to control the ultrasound probe to obtain ultrasound image frames during a blind sweep protocol on a patent with a pregnancy, and provide the ultrasound image frames as an input to a deep learning network trained to detect fetal anatomies. The processor is also configured to generate, as an output of the deep learning network, detections of the fetal anatomies within the ultrasound image frames, generate an anatomical map based on the fetal anatomies, determine that the pregnancy is a multiple gestation based on the anatomical map, and display an output representative of the determination that the pregnancy is a multiple gestation.

The subject-matter of the present disclosure relates to a mobile radiography apparatus (100), comprising a first set of sensors (116), comprising at least one sensor, configured to receive a signal emitted by a beacon device (200), a controller (118) configured control a linear movement of the mobile radiography apparatus based on the received signal, a second set of sensors (120), comprising at least one sensor, configured to detect a position of the beacon device, wherein the controller is configured to control a rotational movement of the mobile radiography apparatus to adjust an alignment of the mobile radiography apparatus with respect to the beacon device, based on the detected position of the beacon device.

Methods and systems for magnetic resonance imaging at least a part of a body are presented. The method comprises: acquiring signals by subjecting the portion of the body (10) to first and second contrast encoding elements within one repetition time (TR) of a sequence, wherein the first contrast encoding elements comprise chemical exchange saturation transfer (CEST) information encoding and the second contrast encoding elements comprise diffusion information encoding; deriving CEST information based on the signals associated with the first contrast encoding elements; deriving diffusion information based on the signals associated with the second encoding elements; and outputting the CEST information and the diffusion information. The system comprises a computational system (15) configured to perform the method.

Disclosed is a method (300) for parameter setting for a medical imaging procedure. The method comprises: controlling (S1) a motorized patient support (120) to apply an excitation to patient body tissue; monitoring (S2) a response of the patient body tissue to the applied excitation; determining (S3) composition of the patient body tissue based on the monitored response; and determining (S4) values of one or more parameters for use in conjunction with the medical imaging procedure based on the determined composition. The method facilitates resource efficient body composition estimation for parameter setting during medical imaging procedures.

A method for determining positions of calipers includes performing calibration to identify preferred placements of calipers in template ultrasound images. The method includes receiving ultrasound image of a patient, inputting the ultrasound image to a first machine learning model for automatically detecting an anatomical view of an anatomy and cropping the ultrasound image to show an ROI, retrieving a matching template ultrasound image that matches the detected anatomical view and preferred placements of the calipers, encoding the preferred placements to provide position embedding, encoding the matching template ultrasound image to provide a first image embedding, encoding the cropped ultrasound image to provide a second image embedding, inputting the position embedding, the first image embedding, and the second image embedding to a trained second machine learning model for predicting positions of calipers in the cropped ultrasound image. The method includes displaying the predicted positions of the calipers, enabling placement of the calipers at the predicted positions.

Disclosed methods and systems relate to denoising a basis material spectral CT image of an anatomical structure. It involves obtaining a noise estimate image with noise values associated with different parts of the spectral CT image, and a structural image describing the anatomical structure. The noise estimate image is modified based on the structural image. For example, the noise estimate image may indicate the presence of more noise in parts of the spectral CT image that are associated with air, as indicated by the structural image. A denoised spectral CT image is then generated using a spectral image denoising algorithm, based on the spectral CT image and the modified noise estimate image. Accordingly, a denoised spectral CT basis material image that may be used to generate certain spectral results images of higher quality may be provided.

A hair care device with a recirculating air flow. The air flow is directed out of an air outlet and into a hair care volume that lies between two elongate housings of the hair care device. An air inlet receives at least a portion of the air flow through the hair care volume, which portion is recirculated to the air outlet via an air flow generator of the hair care device.

A computer-implemented method of generating responses, e.g. answering questions, using an artificial intelligence, Al, agent, is provided. The method includes receiving a plurality of consecutive statements; and invoking the AI agent to generate, and to output, responses corresponding to the statements. For at least one of the statements, the invoking comprises: combining the statement with one or more earlier statements to provide a combined statement; and generating, and outputting, a response to the combined statement. The generating of the response to the combined statement is performed concurrently with the generating of the response to at least one of the one or more earlier statements.

The invention relates to an actively cooled ultrasound system. The system includes a cable (200) comprising a cable connector (250) with an inlet (215) and outlet (235) for coolant, a heat pump to draw the coolant, and an ultrasound probe (100) with an imaging sensor (120) and an electronic circuit (150). An open-loop cooling circuit circulates the coolant past the electronic circuit to absorb and transport heat away. The cable has a first coolant conduit (210) to transport coolant to the probe and a second coolant conduit (230) to transport coolant back to the connector.

The present application teaches an improved apparatus (500) and method arranged to calculate an opto-electronic transfer function (OETF_PU) which is arranged to transform a linear input color coordinate (L_i) into a non-linear output color coordinate (Y_i_PU), wherein the apparatus comprises a coarse OETF calculation unit (510) arranged to calculate a coarse approximation of the opto-electronic transfer function by calculating for the input color coordinate (L_i) a non-linear intermediate color coordinate (Y_crs), by using a coarse OETF (F_crs) which consists of a number of segments between non-uniform endpoints, wherein in each segment the output color coordinate is defined as a function of the input color coordinate which function is a square root;and wherein the apparatus comprises a fine adjustment unit (511) which is arranged to apply a fine-tuning function (F_ft) to the non-linear intermediate color coordinate (Y_crs) to obtain the output color coordinate