An image processing apparatus includes a hardware processor. The hardware processor is configured to determine a standard frame image in a dynamic image including a plurality of frame images obtained by imaging using radiation a process of a movement in an intentional motion by a subject. The hardware processor is configured to set a plurality of feature points on a predetermined structure in the standard frame image. The hardware processor is configured to generate position information showing a position state of the predetermined structure in the standard frame image from a relation of positions of the plurality of set feature points. The hardware processor is configured to specify in another frame image of the dynamic image positions of the plurality of set feature points and generates the position information of the predetermined structure in the another frame image.

A system and method include identification of a plurality of sub-volumes of an image volume, each of the plurality of sub-volumes of the image volume associated with a respective one of a plurality of vertebra, registration of each of the plurality of sub-volumes of the image volume to one of a plurality of reference sub-volumes associated with a same respective vertebra, input of the registered sub-volumes and the image volume to a trained neural network, reception of an output image volume from the trained neural network, the output image volume labeled to associate voxels with respective vertebra, and display of the output image volume.

Disclosed herein are a method and a system for creating a registered image that integrates the information of CT and CBCT images. With the present method and system, medical practitioners can precisely transform the information of CT image-based treatment plan into the CBCT image so as to accurately control the dosage and location of a radiation therapy. Accordingly, also disclosed herein are methods of treating a cancer in the subject with the aid of the method and/or system of the present disclosure.

A method of medical image registration is to be implemented by a computer device, and includes steps of: obtaining an ultrasound target image corresponding to an area of interest in an ultrasound image; for each of multiple computed tomography (CT) images, obtaining a CT candidate image that corresponds to an area of interest in the CT image; for each of the CT candidate images, calculating a similarity between the CT candidate image and the ultrasound target image; making one of the CT candidate images that corresponds to the greatest similarity among the CT candidate images serve as a CT target image; and performing image registration on the ultrasound target image and the CT target image.

Camera arrays for mediated-reality systems and associated methods and systems are disclosed herein. In some embodiments, a camera array includes a support structure having a center, and a depth sensor mounted to the support structure proximate to the center. The camera array can further include a plurality of cameras mounted to the support structure radially outward from the depth sensor, and a plurality of trackers mounted to the support structure radially outward from the cameras. The cameras are configured to capture image data of a scene, and the trackers are configured to capture positional data of a tool within the scene. The image data and the positional data can be processed to generate a virtual perspective of the scene including a graphical representation of the tool at the determined position.

Apparatus and methods are described including acquiring 3D image data of a targeted vertebra. A processor indicates the targeted vertebra within the 3D image data. A radiopaque element is positioned on the body of the subject with respect to the spine and a radiographic image is acquired. The processor (a) generates a plurality of 2D projections of the targeted vertebra from the 3D image data, (b) for each vertebra that is visible in the radiographic image, identifies if there exists a 2D projection of the targeted vertebra that matches the radiographic image of the vertebra that is visible, and (c) in response, indicates on the 2D radiographic image the vertebra for which a match with a projection of the targeted vertebra was identified, such that a location of the targeted vertebra is identified with respect to the radiopaque element. Other applications are also described.

One or more medical images of a patient are processed by a first neural network model to determine a region-of-interest (ROI) or a cut-off plane. Information from the first neural network model is used to crop the medical images, which serves as input to a second neural network model. The second neural network model processes the cropped medical images to determine contours of anatomical structures in the medical images of the patient. Each of the first and second neural network models are deep neural network models. By use of cropped images in the training and inference phases of the second neural network model, contours are produced with sharp edges or flat surfaces.

Systems and methods are disclosed whereby a surface detection system is employed to obtain intraoperative surface data characterizing an exposed surface of the spine. In some embodiments, this intraoperative surface data is registered to segmented surface data obtained from volumetric data of the spine in order to assess the intraoperative orientation of the spine and provide feedback associated with the intraoperative orientation of the spine. The feedback may characterize the intraoperative spinal orientation as a change relative to the preoperative orientation.

The present disclosure involves a reference line determination method and system. In a process of determining a reference line, a plurality of original images containing a first spatial position information are obtained. According to the plurality of original images, a composite image containing a second spatial position information is further determined. After a composition relationship between a plurality of original images was determined, a reference line is determined on the composite image according to the spatial position information.

The present disclosure provides a computer-implemented method, a device, and a computer program product for radiographic bone mineral density (BMD) estimation. The method includes receiving a plain radiograph, detecting landmarks for a bone structure included in the plain radiograph, extracting an ROI from the plain radiograph based on the detected landmarks, estimating the BMD for the ROI extracted from the plain radiograph by using a deep neural network.