Systems and methods for identifying anatomically relevant blood flow characteristics in a patient

A system for identifying anatomically relevant blood flow characteristics in a patient, the system comprising:

• at least one data storage device storing instructions for identifying anatomically relevant blood flow characteristics in a patient;
• and
• at least one processor configured to execute the instructions to perform a method comprising:
• receiving, in an electronic storage medium, a first patient-specific representation of at least a portion of vasculature of the patient having a lesion at one or more points, the first patient-specific representation including a plurality of features of the at least a portion of the vasculature, the plurality of features including one or more of an anatomical, physiological, or geometrical feature, wherein the plurality of features includes a bifurcation;
• determining a location of a diseased region in the vasculature of the patient, by:
• receiving one or more observed lumen measurements of the vasculature of the patient, each observed lumen measurement corresponding to a respective position along a blood vessel of the at least portion of vasculature of the patient based on the first patient-specific representation;
• predicting a second patient-specific representation of at least a portion of the vasculature of the patient in a healthy condition by inputting the plurality of features of the first patient-specific representation into a trained machine learning algorithm, wherein:
• the second patient-specific representation includes one or more healthy lumen measurements of the vasculature of the patient;
• and
• each healthy lumen measurement in the second patient-specific representation corresponds to a respective one of the observed lumen measurements in the first patient-specific representation;
• generating a respective lumen narrowing score for each position along the blood vessel based on a comparison of each observed lumen measurement and the corresponding healthy lumen measurement;
• determining a location of the lesion along the blood vessel based on the generated one or more lumen narrowing scores;
• determining, based on the generated one or more lumen narrowing scores, a distal bound of the lesion;
• automatically determining that the bifurcation is within a first predetermined distance beyond the determined distal bound in a distal direction;
• automatically determining a first distal position on the first vessel that is the first predetermined distance beyond the determined distal bound;
• automatically determining a second distal position on the second vessel that is the first predetermined distance beyond the determined distal bound;
• and
• generating a visualization of at least the diseased region, the visualization including:
• the first vessel and the second vessel;
• a first distal pin displayed on the first vessel at the determined first distal position;
• and
• a second distal pin displayed on the second vessel at the determined second distal position.

A computer-implemented method of identifying anatomically relevant blood flow characteristics in a patient, the method comprising:

• receiving, in an electronic storage medium, a first patient-specific representation of at least a portion of vasculature of the patient having a first lesion at a first location and a second lesion at a second location, the first patient-specific representation including a plurality of features of the at least a portion of the vasculature, the plurality of features including one or more of an anatomical, physiological, or geometrical feature;
• determining a location of a diseased region in the vasculature of the patient, by:
• receiving one or more observed lumen measurements of the vasculature of the patient, each observed lumen measurement corresponding to a respective position along a blood vessel of the at least portion of vasculature of the patient based on the first patient-specific representation;
• predicting a second patient-specific representation of at least a portion of the vasculature of the patient in a healthy condition by inputting the plurality of features of the first patient-specific representation into a trained machine learning algorithm, wherein:
• the second patient-specific representation includes one or more healthy lumen measurements of the vasculature of the patient;
• and
• each healthy lumen measurement in the second patient-specific representation corresponds to a respective one of the observed lumen measurements in the first patient-specific representation;
• generating a respective lumen narrowing score for each position along the blood vessel based on a comparison of each observed lumen measurement and the corresponding healthy lumen measurement;
• determining, based on the generated one or more lumen narrowing scores, a first distal bound of the first lesion;
• determining, based on the generated one or more lumen narrowing scores, a first proximal bound of the first lesion;
• determining, based on the generated one or more lumen narrowing scores, a second distal bound of the second lesion;
• automatically determining that the first proximal bound is within a first predetermined distance of the second distal bound;
• automatically determining that the diseased region includes the first and second lesions;
• automatically determining a pin position in the diseased region based on the first proximal bound and the second distal bound;
• and
• generating a visualization of at least the disease region, wherein the visualization includes the first lesion, the second lesion, and at least one pin provided at the determined pin position.