Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating data specifying a three-dimensional mesh of an object using an auto-regressive neural network.

Abstract: A method and system for composite molding, and the method includes additively manufacturing a mold shell that includes a part forming cavity. The method includes filling at least a portion of the mold shell with a strengthening agent. The method includes hardening the strengthening agent to form a mold, and laying-up a composite material on the part forming cavity. The method includes curing the mold to form a composite part.

Abstract: Assays for compounds having a desired biological activity against one but not both of a pair of microorganism species utilize separately detectable labels. The first and second microorganisms are different from each other, the first microorganisms exhibit a first detectable label, the second microorganisms exhibit a second detectable label different from the first detectable label, and neither of the labels interferes with detection of the other label. In various embodiments, the first and second microorganisms are incubated with a candidate compound or a producer thereof. The different labels permit isolation of microorganisms exhibiting one of the labels but not the other, indicating the desired activity, and the compound responsible for this differential response is isolated.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating data specifying a three-dimensional mesh of an object using an auto-regressive neural network.

Abstract: A method and system for composite molding, and the method includes additively manufacturing a mold shell that includes a part forming cavity. The method includes filling at least a portion of the mold shell with a strengthening agent. The method includes hardening the strengthening agent to form a mold, and laying-up a composite material on the part forming cavity. The method includes curing the mold to form a composite part.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for performing protein design. In one aspect, a method comprises: processing an input characterizing a target protein structure of a target protein using an embedding neural network having a plurality of embedding neural network parameters to generate an embedding of the target protein structure of the target protein; determining a predicted amino acid sequence of the target protein based on the embedding of the target protein structure, comprising: conditioning a generative neural network having a plurality of generative neural network parameters on the embedding of the target protein structure; and generating, by the generative neural network conditioned on the embedding of the target protein structure, a representation of the predicted amino acid sequence of the target protein.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating compressed representations of synthetic images. One of the methods is a method of generating a synthetic image using a generative neural network, and includes: generating, using the generative neural network, a plurality of coefficients that represent the synthetic image after the synthetic image has been encoded using a lossy compression algorithm; and decoding the synthetic image by applying the lossy compression algorithm to the plurality of coefficients.

Abstract: Novel compounds and analogues that exhibit antimicrobial activity—particularly against Gram-negative pathogens—are producible from bacterial isolate Photorhabdus australis DSM 17609 and some related bacterial species of the genus. Pharmaceutical compositions containing the novel compound and its analogues are useful for treating or preventing a bacterial infection. Compounds in accordance herewith include ribosomally produced and post-translationally modified cyclic peptides useful for the treatment, amelioration, and prevention of bacterial infections by Gram-negative pathogens, in addition to other indications.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating a sequence of data elements that includes a respective data element at each position in a sequence of positions. In one aspect, a method includes: for each position after a first position in the sequence of positions: obtaining a current sequence of data element embeddings that includes a respective data element embedding of each data element at a position that precedes the current position, obtaining a sequence of latent embeddings, and processing: (i) the current sequence of data element embeddings, and (ii) the sequence of latent embeddings, using a neural network to generate the data element at the current position. The neural network includes a sequence of neural network blocks including: (i) a cross-attention block, (ii) one or more self-attention blocks, and (iii) an output block.

Abstract: An automatic protocolling system and methods involving a processor operable by way of a set of executable instructions storable in relation to a nontransient memory device, the set of executable instructions configuring the processor to: receive information relating to an initial protocol comprising an initial ordering of a plurality of sequences, the information comprising data relating to an interaction extent value of at least one of an imaging system and a patient as a function of time corresponding to each sequence in the plurality of sequences, the data relating to a time-integrated effect of each sequence in the plurality of sequences; and dynamically determine an alternative protocol comprising an alternative ordering of the plurality of sequences based on the time-integrated effect, whereby an alternative protocol is provided.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating data specifying a three-dimensional mesh of an object using an auto-regressive neural network.

Abstract: Methods for correcting a non-uniform power response of a radiofrequency (“RF”) transmit coil used in magnetic resonance imaging (“MRI”) are described. Transmit power response data for an RF transmit coil are processed to compute RF amplitude scaling factors for the RF transmit coil as a function of transmit frequency offset. The RF amplitude scaling factors can be used to correct transmitted RF power, and thus flip angle, to be more uniform over a range of transmit frequency offsets, as may be encountered when imaging with lower field MRI systems or MRI systems with high strength or asymmetric gradients.

Abstract: A flow conditioning device includes a body portion having first and second ends and an interior surface defining a channel extending from the first end to the second end, and one or more flow conditioning elements disposed within the channel. Each of the one or more flow conditioning elements is integrally formed with the interior surface of the body portion, and may be a respective flow straightening tube, a flow straightening vane, a hole array plate, a turning vane, a swirling vane, a helical ridge formed along a respective longitudinal segment of the interior surface, or another configuration. The body portion and the one or more flow conditioning elements may be formed by an additive manufacturing process, and may optionally be made of the same material.

Abstract: An asymmetric electromagnet system, method, and method of producing an asymmetric electromagnet system, wherein the asymmetric electromagnet system is for generating an imaging magnetic field in an imaging region with an imaging isocentre, the imaging region being asymmetrically positioned within a gradient coil bore inside a magnetic resonance imaging (MRI) system during imaging, the electromagnet assembly comprising: an asymmetric gradient coil configured to generate a gradient field in the asymmetrically positioned imaging region, at least one gradient axis having the gradient field with a constant offset component such that the position at which the gradient field passes through zero is offset with respect to the imaging isocentre of the asymmetrically positioned imaging region.

Abstract: An underground tunneling detection system and method includes a tunneling detection control unit that receives one or more sound detection signals including a sound signature output by a component that is underground and at or proximate to a location. The tunneling detection control unit compares the sound signature to sound signature data. The tunneling detection control unit determines that the sound signature is non-tunneling activity in response to the sound signature matching a non-tunneling portion of the sound signature data. The tunneling detection control unit determines that the sound signature is tunneling activity in response to the sound signature matching a tunneling portion of the sound signature data. The tunneling detection control unit determines a similarity metric between the sound signature and one or both of the non-tunneling portion of the sound signature data or the tunneling portion of the sound signature data in response to the sound signature differing from the sound signature data.

Abstract: Methods for correcting a non-uniform power response of a radiofrequency (“RF”) transmit coil used in magnetic resonance imaging (“MRI”) are described. Transmit power response data for an RF transmit coil are processed to compute RF amplitude scaling factors for the RF transmit coil as a function of transmit frequency offset. The RF amplitude scaling factors can be used to correct transmitted RF power, and thus flip angle, to be more uniform over a range of transmit frequency offsets, as may be encountered when imaging with lower field MRI systems or MRI systems with high strength or asymmetric gradients.

Abstract: Methods for correcting a non-uniform power response of a radiofrequency (“RF”) transmit coil used in magnetic resonance imaging (“MRI”) are described. Transmit power response data for an RF transmit coil are processed to compute RF amplitude scaling factors for the RF transmit coil as a function of transmit frequency offset. The RF amplitude scaling factors can be used to correct transmitted RF power, and thus flip angle, to be more uniform over a range of transmit frequency offsets, as may be encountered when imaging with lower field MRI systems or MRI systems with high strength or asymmetric gradients.

Abstract: The present disclosure provides a method and system for correcting errors caused by non-linearities in a gradient field profile of a gradient coil in a magnetic resonance imaging (MRI) system. The method includes obtaining a non-linearity tensor at each voxel within the imaging space using a computer model of the gradient coil; correcting motion sensitive encoding using the non-linearity tensor; and generating a corrected image using the corrected motion sensitive encoding.

Abstract: The present disclosure provides a method and system of magnetic resonance imaging using a constrained gradient waveform. The constrained gradient waveform is designed to have only predetermined frequencies, for example excluding one or more identified resonant frequencies associated with the MRI system. The application of such a constrained gradient waveform during imaging may aid in reducing noise, vibration and/or heating of the MRI system during imaging.

Abstract: Some implementations provide a MRI system configured to: access data encoding an input gradient waveform that would otherwise be used on a gradient sub-system of the MRI system to generate a gradient that corresponds to perturbations to the substantially uniform magnetic field; access data encoding a forward impulse response function and an inverse impulse response function, both characterizing a gradient generated from a target impulse input; pre-emphasizing the input gradient waveform by using both the forward impulse response function and the reverse impulse response function; and drive the gradient sub-system using the pre-emphasized gradient waveform such that distortions to the gradient caused by eddy currents within the gradient sub-system are substantially removed while radio-frequency (RF) samples for reconstructing an MRI image are being acquired from a grid that is substantially identical to when gradient sub-system is driven using the input gradient waveform.