Back to the list
Congress: ECR25
Poster Number: C-11699
Type: Poster: EPOS Radiologist (scientific)
Authorblock: I. Ayx, L. Lichti, S. Büttner, T. Papavassiliu, K. Sopova, S. O. Schönberg, M. Kuru, C. A. Marschner; Mannheim/DE
Disclosures:
Isabelle Ayx: Nothing to disclose
Lena Lichti: Nothing to disclose
Sylvia Büttner: Nothing to disclose
Theano Papavassiliu: Nothing to disclose
Kateryna Sopova: Nothing to disclose
Stefan Oswald Schönberg: Nothing to disclose
Mustafa Kuru: Nothing to disclose
Constantin Arndt Marschner: Nothing to disclose
Keywords: Artificial Intelligence, Cardiac, CT-Angiography, Diagnostic procedure, Arteriosclerosis
Results

Patients were divided into CAD RADS 3 (9 patients, 6 male, 3 female, mean age 70 years, median Agatston Score 192.8) and CAD RADS 4a (19 patients, 19 male, 0 female, mean age 67 years, median Agatston Score 505.15). Revascularization was only performed in the CAD RADS 4a group, mean Delta CT-FR differed significantly between both groups (p = 0.0104).

 

There was a significant correlation between the stenosis degree in CCTA and ICA (p = 0.0043), although a trend of overestimating the stenosis degree was seen in CCTA. Stent placement was recorded in 10 cases and the stenosis degree in CCTA was the relevant feature for stent placement (p = 0.0095). The Odds Ratio of 7.6 outlined that patients with CAD RADS 4a have a 7.6 times higher risk of stent placement than patients with CAD RADS 3 (Area under the curve (AUC) = 71.3%).

 

A statistical comparison of CT-FFR and ICA was conducted for the most severe stenosis in each patient, as this lesion guided the diagnostic approach at the patient level. For the CAD RADS 3 group no revascularization was performed, hence negative predictive value was 100 %. In 7 out of 9 patients (77.8 %) no pathological Delta CT-FFR was detected, in 2 cases (22.2 %) Delta CT-FFR was pathological although revascularization was not necessary. Diagnostic accuracy of Delta CT-FFR in the CAD RADS 3 group was hence 77.8 %.

 

In the CAD RADS 4a group, in all patients with stent placement, Delta CT-FFR was pathological. Figure 2 offers an example cases of a CAD RADS 4a patient with pathological Delta CT-FFR and revascularization. In 5 patients (26.3 %) revascularization wasn’t necessary despite pathological Delta CT-FFR. 4 patients (21.1 %) could have been prevented from ICA due to not pathological Delta CT-FFR. Specificity in the CAD RADS 4a group was 44.4 %, sensitivity 100 %, positive predictive value 66.7 %, negative predictive value 100 % and diagnostic accuracy 73.7 %.

 

Summarizing both groups, Delta CT-FFR was considered pathological in all patients, who underwent revascularization. No patients with normal Delta CT-FFR required stent implantation. In 39.29 % of the cases, calculation of Delta CT-FFR could have prevented patients from undergoing unnecessary ICA. The positive predictive value of Delta CT-FFR was 58.8 %, negative predictive value 100 %, and diagnostic accuracy 0.75 (Figure 3).

 

Additionally, the MACE rate was assessed. None of the patients with a normal Delta CT-FFR experienced a MACE within 6 months. Among all 28 patients, 3 (10.7%) had a MACE, with 3 out of 17 patients (17.6%) exhibiting pathological Delta CT-FFR. Notably, 1 of these 3 patients (66.7%) underwent revascularization during the initial ICA.
GALLERY