Classification of medical images using machine learning to account for body orientation

A computer system for identifying a medical condition in a patient, the computer system comprising:

• one or more computer processors;
• one or more computer readable storage media;
• 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 program instructions comprising instructions to:
• generate, using a trained machine learning image generator, a set of training images, wherein the set of training images is generated based on three-dimensional imaging data from a plurality of patients, wherein each training image is based on a two-dimensional projection of the three-dimensional imaging data of a particular patient, and wherein each training image is labeled with a projection angle of the corresponding two-dimensional projection;
• train, using the set of training images, a machine learning image classifier model to identify patient rotation angles in x-ray images;
• process a set of x-ray images with the machine learning image classifier model to identify a patient rotation angle for each x-ray image, wherein each x-ray image is labeled with a disease state;
• train a machine learning medical condition classifier model to identify a medical condition, wherein the machine learning medical condition classifier model is trained using the set of x-ray images labeled with the medical condition state and the patient rotation angle determined by the image classifier model;
• and
• apply the machine learning medical condition classifier model to determine an indication of the medical condition in an input x-ray image acquired from a patient.

A computer program product for identifying a medical condition in a patient, the computer program product comprising one or more computer readable storage media collectively having program instructions embodied therewith, the program instructions executable by a computer to cause the computer to:

• generate, using a trained machine learning image generator, a set of training images, wherein the set of training images is generated based on three-dimensional imaging data from a plurality of patients, wherein each training image is based on a two-dimensional projection of the three-dimensional imaging data of a particular patient, and wherein each training image is labeled with a projection angle of the corresponding two-dimensional projection;
• train, using the set of training images, a machine learning image classifier model to identify patient rotation angles in x-ray images;
• process a set of x-ray images with the machine learning image classifier model to identify a patient rotation angle for each x-ray image, wherein each x-ray image is labeled with a disease state;
• train a machine learning medical condition classifier model to identify a medical condition, wherein the machine learning medical condition classifier model is trained using the set of x-ray images labeled with the medical condition state and the patient rotation angle determined by the image classifier model;
• and
• apply the machine learning medical condition classifier model to determine an indication of the medical condition in an input x-ray image acquired from a patient.