Coronary CT Angiography

Value Proposition

Coronary CT Angiography offers opportunities to minimize patient risk compared to invasive coronary angiography in efforts to diagnosis coronary artery disease. While improvements in image quality have been made recently, AI that delivers metrics related to CTA would be promising for clinician efficiency and patient outcomes. Currently CTA faces some challenges to invasive procedures such as lower temporal resolution, greater frequency of motion artifacts, and higher false-positives from severe calcification.

Narratives

A 55 year old woman presents to the emergency department with chest pain. Initial blood work is negative for a myocardial infarction (heart attack). She has a history of high cholesterol putting her at an intermediate risk for an acute myocardial infarction. Her doctor orders an ECG that appears normal. A coronary CTA is ordered to look for coronary artery disease.
A 60 year old man presents to his primary care physician with a 3 week history of chest pain. Blood work and EKG are negative. The nuclear stress test is equivocal. His doctor orders a coronary CTA to look for coronary artery disease.

Workflow Description

The image is obtained from the CT scanner, sent to the AI algorithm and to the PACS. The CT images are post-processed by the algorithm. The algorithm measures the minimum diameter of the coronary arteries. The post-processed images are sent to the PACS.

Inputs

DICOM Study

Primary Outputs

Percent Stenosis of the Coronary Arteries

Maximum Percent Stenosis of the left main coronary artery

Maximum percent stenosis of the left anterior descending coronary artery

Minimum percent stenosis of the diagonal branches

Minimum percent stenosis of the ramus branch

Minimum percent stenosis of the left circumflex artery

Minimum percent stenosis of the obtuse marginal branch

Minimum percent stenosis of the right coronary artery

Minimum percent stenosis of the PDA

Minimum percent stenosis of the posterior lateral branch

• The algorithm could give CAD-RADs categories
  
  

No known related public datasets at this time,  please alert us if you know of any.