Automatic detection of vertebral dislocations

A computer-implemented method for automatic detection of vertebral dislocations, the computer-implemented method comprising:

• receiving, by one of more computer processors, one or more sets of training data, wherein the one or more sets of training data include an ordered sequence of radiographic image patches of vertebrae and intervertebral spaces and a location data for each vertebra and each intervertebral space in the radiographic image patches;
• training, by the one or more computer processors, one or more location deep learning models to detect a location of each vertebra of one or more vertebrae and each intervertebral space of one or more intervertebral spaces from the one or more sets of training data,
• training, by the one or more computer processors, one or more classification deep learning models to classify an ordered sequence of image patches to identify one or more vertebral anomalies of the one or more vertebrae and the one or more intervertebral spaces from the one or more sets of training data,
• receiving, by the one or more computer processors, one or more computed tomography (CT) image files;
• determining, by the one or more computer processors, a vertebral body center and an intervertebral disc space for each vertebra of one or more vertebrae in a field of view of the one or more CT image files;
• extracting, by the one or more computer processors, one or more parameters including at least one of an intervertebral disc size, an intervertebral disc volume, and a Hounsfield unit value for each vertebra of one or more vertebrae in the field of view of the one or more CT image files;
• identifying, by the one or more computer processors, one or more CT subvolumes, wherein each CT subvolume contains a vertebra;
• creating, by the one or more computer processors, one or more two-dimensional projections from the one or more CT subvolumes, wherein the one or more two-dimensional projections are created using volume rendering techniques;
• and
• applying, by the one or more computer processors, the one or more location deep learning models and the one or more classification deep learning models to the one or more radiographic image files to create a condition assessment.

A computer system for automatic detection of vertebral dislocations, the computer system comprising:

• one or more computer processors;
• one or more computer readable storage media;
• and
• program instructions stored on the one or more computer readable storage media for execution by at least one of the one or more computer processors, the stored program instructions including instructions to:
• receive one or more sets of training data, wherein the one or more sets of training data include an ordered sequence of radiographic image patches of vertebrae and intervertebral spaces and a location data for each vertebra and each intervertebral space in the radiographic image patches;
• train one or more location deep learning models to detect a location of each vertebra of one or more vertebrae and each intervertebral space of one or more intervertebral spaces from the one or more sets of training data,
• train one or more classification deep learning models to classify an ordered sequence of image patches to identify one or more vertebral anomalies of the one or more vertebrae and the one or more intervertebral spaces from the one or more sets of training data, and
• responsive to receiving one or more radiographic image files, apply the one or more location deep learning models and the one or more classification deep learning models to the one or more radiographic image files to create a condition assessment.