Summarization of clinical documents with end points thereof

A non-transitory computer-readable storage medium carrying instructions which controls one or more processors to:

• determine relevant data and discard irrelevant data in one or more clinical report objects for a subject;
• generate a machine learning model from training data that includes validated end points and clinical report objects, wherein the machine learning model is configured to:
• predict an end point using the one or more clinical report objects for a subject, wherein the machine learning model inputs the one or more clinical report objects and outputs the predicted end point according to phrases or n-grams in the one or more clinical report objects,
• wherein the machine learning model is based on a convolutional approach using minimal feature engineering, and
• wherein the convolutional neural network applies at least one or more analytical techniques including deep learning, a random forest classifier, a support vector machine, and logistic regression;
• and
• visualize the predicted end point, wherein the predicted end point includes at least one of a repeated imaging study within a first predetermined time frame, an oncological malignancy diagnosed within a second predetermined time frame, a life time expectancy within a third predetermined time frame, a qualification for a screening program within a fourth predetermined time frame, an occurrence of a major cardiac event within a fifth predetermined time frame, a hospital cost of an episode of care greater than a predetermined amount within a sixth predetermined time frame, and a hospital readmission within a seventh predetermined time frame.