Methods and systems for analyzing and reporting medical images

A method comprising:

• retrieving, by a computer system, scan data of a predetermined region of a patient's body, the scan data comprising at least one image of the predetermined region;
• automatically analyzing, using the computer system, the scan data to identify an abnormality relating to the patient's condition using an image processing algorithm;
• automatically performing, using the computer system, quantitative measurements relating to the abnormality using image analysis, wherein the quantitative measurements are associated with at least one of a distance, a cross-sectional area, and a volume relating to the patient's condition;
• automatically identifying, using the computer system, a representative image from the scan data based on the identified abnormality and the quantitative measurements relating to the abnormality;
• automatically generating, using the computer system, an analysis report including at least one of the representative image and the quantitative measurements associated with the representative image;
• retrieving, using the computer system, a historical analysis report of the predetermined region of the patient's body, wherein the historical analysis report includes at least one of a historical representative image and quantitative measurements associated with the historical representative image;
• automatically comparing, using the computer system, the quantitative measurements relating to the abnormality in the analysis report with the quantitative measurements relating to the abnormality in the historical analysis report to identify a change in the patient's condition;
• determining, using a machine learning model, a change in diagnosis associated with the patient based at least in part on the identified change in the patient's condition, the change in diagnosis being indicative of a potential problem in a historical diagnosis or a progression of the patient's condition, wherein the machine learning model was generated by:
• receiving a plurality of final reports associated with the physician and a corresponding plurality of preliminary reports generated by an initial machine learning model;
• comparing the plurality of final reports by the physician with the corresponding plurality of preliminary reports to adjust the initial machine learning model to more closely match the plurality of final reports;
• and
• further refining the initial machine learning model based at least in part on patient outcome information relating to patient responsiveness to treatment, the treatment being associated with conditions identified in the plurality of preliminary reports generated by the initial machine learning model;
• and
• presenting, using the computer system, a preliminary report associated with the change in diagnosis to a physician for review, wherein the preliminary report associated with the change in diagnosis includes at least one of (i) the representative image and historical representative image, (ii) the quantitative measurements relating to the abnormality in the analysis report and the quantitative measurements relating to the abnormality in the historical analysis report, and (iii) the change in diagnosis determined by the machine learning model.