Noninvasive assessment of microvascular dysfunction

A method for assessing coronary microvascular dysfunction of a patient, the method comprising:

• acquiring a first series of images of a region of interest within the patient;
• determining a first angiography-derived index of microcirculatory resistance from the series of images and a measured first aortic pressure of the patient, wherein determining the imaging-derived index of microcirculatory resistance from the series of images comprises:
• measuring the first aortic pressure of the patient;
• determining an angiography-derived fractional flow reserve from the series of images;
• determining, from the series of images, a transit time of a marker to travel between first and second reference points;
• and
• calculating a product of a ratio of the number of frames necessary for the contrast dye to travel from the guiding catheter to the distal reference to a frame rate of the imager, the first aortic pressure, and the angiography-derived fractional flow reserve determined from the series of images;
• assigning a first clinical parameter to the patient via a first predictive model trained on a set of training data representing the first clinical parameter and receiving at least the first angiography-derived index of microcirculatory resistance as an input and providing the first clinical parameter as an output, wherein the first clinical parameter is a categorical parameter that can assume a first value representing a presence of microvascular dysfunction, a second value representing an absence of microvascular dysfunction, and a third value representing an indeterminate result;
• and
• performing the following steps if the first clinical parameter assumes the third value:
• providing a vasodilator to the patient;
• acquiring a second series of images of the region of interest;
• determining a second angiography-derived index of microcirculatory resistance from the second series of images and a second measured aortic pressure of the patient;
• and
• assigning a second clinical parameter to the patient via a second predictive model trained on a set of training data representing the second clinical parameter and receiving at least the second angiography-derived index of microcirculatory resistance as an input and providing the second clinical parameter as an output, wherein the second clinical parameter is a categorical parameter that can assume a first value representing the presence of microvascular dysfunction and a second value representing the absence of microvascular dysfunction.