“Hands free” artificial intelligence (AI) based angiography via imaging-derived fractional flow reserve (FFR) with the AutocathFFR was found to match the performance of wire-based FFR measurements in a new 304-patient pivotal single-arm study. The results were reported in a late-breaking session Sunday at Cardiovascular Research Technologies (CRT) 2022 in Washington, D.C., by Hector M. Garcia-Garcia, MD, PhD, MedStar Washington Hospital Center. MedHub’s AutocathFFR is an automated non-invasive FFR assessment software device that utilizes digital angiography images. The aim is to replace “risky and costly invasive procedures” with AI image-based tools, MedHub said in in a press statement on the results. Hemodynamically significant coronary artery disease (CAD) must be thoroughly assessed as part of intervention decision-making to determine stenosis diameter and hemodynamic significance. Although FFR is accurate – and is the most commonly used method for these analyses – the method is invasive, carries the risk of complications and is time-consuming, said Garcia-Garcia. Angio-derived FFR technologies do not carry the same risk of complications because they do not require a wire, Garcia-Garcia added, “and it can be calculated out of routinely taken angio views.” The study used the prospective analysis of retrospective data to compare the gold-standard invasive FFR with the non-invasive AutocathFFR measurements in 304 patients with known or suspected CAD who had undergone invasive coronary angiography to assess a non-culprit narrowing in at least one coronary artery. The cohort was aged 18 years or older and included those with stable angina pectoris, unstable angina pectoris or non-ST-segment elevation myocardial infarction (NSTEMI). At baseline, patients’ mean age was 63.7±10 years, the majority were male (78.8%), mean body mass index was 28.99±5.09 kg/m2, and 47.8% had diabetes mellitus. Those with vessel diameter of less than 2 mm, chronic total occlusion (CTO) in the target vessel or providing collaterals to a chronically occluded vessel, prior percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG) were excluded. FFR was measured according to current guidelines, with FFR ≤0.80 in invasive procedures scored as “positive” and FFR >0.8 considered “negative.” In the overall population (304 vessels), Autocath FFR accuracy was 93.7%. In the “gray zone” (136 vessels between 0.77 and 0.87), accuracy was 94.1%. The overall mean invasive FFR was 0.8575, and the AutoFFR was 0.8521 (a mean difference of 0.00539, p = 0.159). The diagnostic performance of AutocathFFR in terms of sensitivity was 0.913 and specificity was 0.945. The positive predictive value was 0.829, while the negative predictive value was 0.974. The overall accuracy was 0.938 and area under the curve (AUC) was 0.93 (95% confidence interval [CI]: 0.89-0.97). The correlation coefficient was 0.53 (95% CI: 0.44-0.61) with a p-value < 0.001. Similar correlations were observed when breaking down the data by X-ray cath lab equipment. In his concluding slides, Garcia-Garcia said the AutocathFFR device “has an excellent diagnostic performance against invasive FFR,” with 100% reproducible results. “Because it is a hands-free analysis, its integration into the catheterization laboratory enables an automated lesion and angio-derived FFR analysis,” he added, stressing that it requires only two – “not necessarily orthogonal” – views per lesion, and does not require adenosine injection.