Abstract: A system for determining subject orientation includes a medical imaging sensor configured to capture medical imagery of an imaging subject. The system also includes one or more interoceptive sensors, which are configured to obtain interoceptive sensor data indicative of an orientation associated with the system.

Abstract: A method for co-registering imagery of different modalities includes accessing a first image captured according to a first imaging modality that depicts a plurality of physical markers. The plurality of physical markers is present within a scene represented in the first image. The method includes accessing a second image captured according to a second imaging modality that captures at least a portion of the scene including the plurality of physical markers. The method further includes generating a transformation matrix using the first image data and the second image data and generating an aligned image by applying the transformation matrix to the first image or the second image. The aligned image is combinable with the first image or the second image to generate co-registered image output. The co-registered image output depicts the portion of the scene according to the first imaging modality and the second imaging modality.

Abstract: A method for identifying a misidentified study can utilize a set of photographs captured at substantially the same time as a corresponding set of medical images. The method can include determining similarities between the photographs through machine learning models and determining that a misidentified study exists when the similarity between the photographs fails to satisfy a threshold similarity.

Abstract: A method for preventing wrong-patient errors includes receiving a selection of a current imaging subject. The current imaging subject is selected for a current image acquisition session comprising capturing one or more current images of the current imaging subject utilizing at least a first image sensor system of a first imaging modality. The method includes accessing one or more previous images of a previous imaging subject. The one or more previous images depict the previous imaging subject according to at least a second imaging modality that is different from the first imaging modality. The method includes presenting the one or more previous images on a display system and, in response to determining that the previous imaging subject matches the current imaging subject based upon the one or more previous images, performing the current image acquisition session.

Abstract: Methods for identifying a disease state in a patient and/or for treating a patient having the identified disease state are disclosed and can be based on characteristics identified through machine learning models such as deep learning convolutional neural networks and that are associated with video recordings, audio recordings, infrared images, photographs, and/or radiologic patient images.

Abstract: Methods for co-registering images of different imaging modalities include obtaining a photograph and a medical image, maximizing mutual information between the photograph and the medical image, automatically locating features common to both the photograph and the medical image based on maximized mutual information between the photograph and the medical image, and automatically highlighting the located features within the photograph or medical image.

Abstract: A method for preventing wrong-patient errors includes receiving a selection of a current imaging subject. The current imaging subject is selected for a current image acquisition session comprising capturing one or more current images of the current imaging subject utilizing at least a first image sensor system of a first imaging modality. The method includes accessing one or more previous images of a previous imaging subject. The one or more previous images depict the previous imaging subject according to at least a second imaging modality that is different from the first imaging modality. The method includes presenting the one or more previous images on a display system and, in response to determining that the previous imaging subject matches the current imaging subject based upon the one or more previous images, performing the current image acquisition session.

Abstract: Methods for identifying a disease state in a patient and/or for treating a patient having the identified disease state are disclosed and can be based on characteristics identified through machine learning models such as deep learning convolutional neural networks and that are associated with video recordings, audio recordings, infrared images, photographs, and/or radiologic patient images.

Abstract: A method for identifying a misidentified study can utilize a set of photographs captured at substantially the same time as a corresponding set of medical images. The method can include determining similarities between the photographs through machine learning models and determining that a misidentified study exists when the similarity between the photographs fails to satisfy a threshold similarity.

Abstract: Patient misidentification errors in medical imaging can result in serious consequences, such as the misdiagnosis of a disease state or the application of an inappropriate treatment regimen. Systems, methods, and apparatuses disclosed herein can properly and consistently identify, adjust, and/or correlate radiologic images with the correct patient. Systems and methods for automated and robust image capture are also provided. Methods for identifying a disease state in a patient and/or for treating a patient having the identified disease state are disclosed and can be based on characteristics identified through deep learning convolutional neural networks and that are associated with photographic and radiologic patient images.

Abstract: Patient misidentification errors in medical imaging can result in serious consequences, such as the misdiagnosis of a disease state or the application of an inappropriate treatment regimen. Systems, methods, and apparatuses disclosed herein can properly and consistently identify, adjust, and/or correlate radiologic images with the correct patient. Systems and methods for automated and robust image capture are also provided. Methods for identifying a disease state in a patient and/or for treating a patient having the identified disease state are disclosed and can be based on characteristics identified through deep learning convolutional neural networks and that are associated with photographic and radiologic patient images.

Abstract: Systems, methods, and computer-readable storage media relate to generate an image series that includes a patient image and a medical image. The patient image and the medical image may be associated based on identification information.

Abstract: An ultrasonic imaging system acquires echo signals from an object being imaged such as a moving coronary artery and the cross-correlation between echo signals is employed as an objective measure of relative object location. The method is used in a prescan procedure to determine an optimal gating window to acquire image data during a cardiac gated scan, and it is used during the scan as a real time gating signal.

Abstract: An x-ray CT system performs a scan by acquiring projection views from which an image is reconstructed. In a prospective embodiment, the correlation of adjacent views is calculated as the scan is performed and is used to detect subject motion as the scan is being performed. In a retrospective embodiment, the correlation of adjacent views is calculated and is used to detect subject motion after the scan is completed. In the first embodiment substitute projection views are acquired by continuing the scan and in the second embodiment redundant projection views acquired during the scan are substituted until the best possible image is produced.

Abstract: An x-ray CT system performs a scan by acquiring projection views from which an image is reconstructed. In a prospective embodiment, the correlation of adjacent views is calculated as the scan is performed and is used to detect subject motion as the scan is being performed. In a retrospective embodiment, the correlation of adjacent views is calculated and is used to detect subject motion after the scan is completed. In the first embodiment substitute projection views are acquired by continuing the scan and in the second embodiment redundant projection views acquired during the scan are substituted until the best possible image is produced.

Abstract: An ultrasonic imaging system acquires echo signals from an object being imaged such as a moving coronary artery and the cross-correlation between echo signals is employed as an objective measure of relative object location. The method is used in a prescan procedure to determine an optimal gating window to acquire image data during a cardiac gated scan, and it is used during the scan as a real time gating signal.