In some embodiments, spinal disc degeneracy is diagnosed according to a decay variance map generated by determining a variance in T2 decay over time within each pixel or pixel subset of an MRI image. A region of interest may be defined as including nucleus pulposus (NP) and annulus fibrosus (AF) areas, and excluding cartilaginous endplate (EP) areas of a disc. A decay variance for a pixel or groups of pixels is calculated by determining ratios between consecutive intensity values of the T2 signal, determining differences between consecutive ratios, and summing the absolute values of the determined differences. Decay variance mapping may be used to diagnose degeneracy in other tissues, such as in joints.

Described herein are various implementations of systems and methods for determining likelihood of a patient favorably responding to a neuromodulation procedure based on a quantitative or objective score or determination based on a plurality of indicators of pain (e.g., chronic low back pain stemming from one or more vertebral bodies or vertebral endplates of a patient). The systems and methods may involve application of artificial intelligence techniques (e.g., trained algorithms, machine learning or deep learning algorithms, and/or trained neural networks).

An image processing apparatus comprising: a processor and a memory connected to or incorporated in the processor, in which the processor acquires an analysis target image in which a plurality of contiguous target objects of the same type appear, receives an input of a marker indicating positions of the target objects in the analysis target image, generates a marker position display map indicating a position of the marker in the analysis target image, inputs the analysis target image and the marker position display map to a semantic segmentation model, and outputs, from the semantic segmentation model, an output image in which the target objects are identified.

A reference part extracting unit extracts at least one reference part that is common in a first image and a second image, from each of the first image and the second image. A first position information deriving unit derives first position information indicating a relative position of at least one abnormal part specified in the first image, relative to the at least one reference part in the first image. A second position information deriving unit derives second position information indicating a relative position of at least one abnormal part specified in the second image, relative to the at least one reference part in the second image. A matching unit associates, on the basis of a difference between the first position information and the second position information, the abnormal part included in the first image and the abnormal part included in the second image with each other.

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.

To provide a technique with which precision of inference can be improved in the case that grayscale images, such as CT images, are handled, there is provided an inference apparatus comprising: an inference section 53 for performing inference using a learned model TM, the learned model TM being generated by learning processing of learning a multi-channel image IMa containing image information on each of three one-channel images (an original image IM1, a histogram-equalized image IM2, and edge-enhanced image IM3), and ground-truth data CD; and a multi-channel image producing section 52 for producing a multi-channel image IMb containing image information on each of three one-channel images (an original image IM10, a histogram-equalized image IM20, and an edge-enhanced image IM30) of a patient, wherein the inference section 53 inputs the multi-channel image IMb to the learned model TM, performs the inference, and outputs an output image IMout as output data.

A method for spinal disorder diagnosis and treatment planning comprising the steps of: imaging a body including vertebral tissue; acquiring data points corresponding to a surface of the body adjacent to the vertebral tissue with a mixed reality holographic display; transmitting the imaging to a computer database; superimposing a first holographic image of the vertebral tissue with a body image including the surface; and displaying the first holographic image and the body image with the mixed reality holographic display. Systems, spinal constructs, implants and surgical instruments are disclosed.

A method and a device for determining a respective position of one or more markers in a 3D image coordinate system are provided. A plurality of image data sets taken from a 3D volume in which an object and one or more markers are disposed. The 3D volume comprises a central volume containing at least a portion of the object and further comprises a peripheral volume adjacent to the central volume and containing the one or more markers. The image data sets have been taken from at least one of different positions and different orientations relative to the object. A first subset comprises image data sets that each includes at least one dedicated marker of the one or more markers and a second subset comprises at least one image data set that does not include the at least one dedicated marker. The method further comprises determining, from the image data sets, a position of the at least one dedicated marker in a 3D image coordinate system of a 3D reconstruction of the central volume with the object portion.

A method of positioning an imaging device relative to a patient, comprising positioning a reference marker adjacent a desired field of scan corresponding to an anatomical element of a patient; and causing an imaging device to align with the reference marker, based on tracking information received from a navigation system, the tracking information corresponding to the reference marker and a navigated tracker disposed on the imaging device.

Methods and systems are provided for whole-body spine labeling. In one embodiment, a method comprises acquiring a non-functional image volume of a spine, acquiring a functional image volume of the spine, determining at least one spine label seed point on a non-functional image volume, automatically labeling the non-functional image volume with a plurality of spine labels based on the at least one spine label seed point, automatically correcting the geometric misalignments and registering the functional image volume, adjusting the plurality of spine labels and propagating the adjusted spine labels to the functional image volume. In this way, the anatomical details of non-functional imaging volumes may be leveraged to improve clinical diagnoses based on functional imaging, such as diffusion weighted imaging (DWI).