Methods and systems for computer-aided diagnosis with deep learning models

A method, comprising:

• training a baseline neural network model on diagnostic decisions of a plurality of radiologists;
• training the baseline neural network model on diagnostic decisions of a subset of the plurality of radiologists corresponding to a geographic region to generate a geographic model;
• training the geographic model on diagnosis decisions of a subset of the subset of the plurality of radiologists corresponding to an institution in the geographic region to generate an institution model;
• training the institution model on diagnostic decisions of a radiologist of the plurality of radiologists, the radiologist associated with the institution in the geographic region, to generate a radiologist model;
• generating, with the radiologist model, a computer-aided diagnosis for a patient;
• receiving feedback from the radiologist regarding the computer-aided diagnosis;
• and
• selectively updating one or more of the baseline neural network model, the geographic model, the institution model, and the radiologist model based on the feedback.

A system, comprising:

• an imaging system configured to acquire a medical image of a subject;
• a display device configured to display the medical image;
• and
• a computing system comprising a processor and a non-transitory memory, the non-transitory memory storing a plurality of neural network models initially trained on diagnostic decisions of a plurality of radiologists, the non-transitory memory configured with instructions that when executed cause the processor to:
• train a first neural network model of the plurality of neural network models based on diagnostic decisions provided by a first radiologist for a first set of medical images acquired via the imaging system;
• train a second neural network model of the plurality of neural network models based on diagnostic decisions provided by a second radiologist for a second set of medical images acquired via the imaging system;
• receive an indication of a radiologist to review the medical image;
• generate a first computer-aided diagnosis with the first neural network model responsive to the indication of the radiologist indicating the first radiologist;
• generate a second computer-aided diagnosis with the second neural network model responsive to the indication of the radiologist indicating the second radiologist;
• and
• display, via the display device, the first computer-aided diagnosis or the second computer-aided diagnosis with the medical image.