Scientific Evidence
As Spimed-AI is based on medical research and development,
we publish our studies in specialized scientific journals.
Assessment of Fractional Flow Reserve from Coronary CT Angiography Using a Deep Learning-Based Algorithm:
A Multicenter Retrospective Study
This multicenter retrospective study evaluated the diagnostic performance
of a deep learning–based algorithm for the non-invasive estimation of fractional flow reserve from coronary CT angiography (CCTA), using invasive physiological measurements as the reference standard. The model demonstrated high diagnostic accuracy for detecting hemodynamically significant coronary artery disease
(per-patient AUC = 0.935; sensitivity = 93.2%; specificity = 93.7%; accuracy = 93.3%).
February 2025
Clinical Cases
Case 1 : AI-Assisted Automatic Extraction of curved MPR coronary CT images
Patient: Female
Age: 42
Background:
Underwent coronary CT angiography (CCTA) for chest discomfort.
Automatic extraction of the 3 main coronary arteries was performed by a new research segmentation tool from CorEx, providing automated CAD-RADS classification and FFR prediction as well.
A subocclusive coronary stenosis from a non calcified plaque was detected on mid RCA confirmed by invasive coronary angiography. Patient was successfully treated by a stent (Dr Patrick Dupouy).
This case illustrates that CorEx software may help detecting automatically obstructive lesions, without needs of any human intervention. AI may then help to democratize the usage of coronary CTA.
Case 2 : Spontaneous LAD dissection revealed by coronary CT angiography
Patient: Female
Age: 46
Background:
No cardiovascular risk factors. Subject underwent coronary CT angiography (CCTA) for acute coronary syndrome
and had blood elevation of troponins: Patient underwent coronary CT angiography (CCTA).
AI-based CorEx software with automated segmentation revealed long LAD stenosis.
Surrounded with diffuse mural thickening, without any calcification, compatible with a coronary wall hematoma. Invasive coronary angiography was subsequently performed confirming moderate stenosis of the proximal LAD, with intra-mural LAD hematoma confirmed by Optical Coherence Tomography (Dr Khalil Mammoudi).
Patient was treated medically with Aspirin and beta-blockers, leading to total recovery and complete regression
of the stenosis on the CCTA performed 3 months later.
Spontaneous coronary dissection is a rare cause of acute coronary syndrome, often affecting women between
30 and 60 years.
CCTA may be useful for diagnosis and follow-up
Case 3 : Coronary CCTA before returning to sport activity
Patient: Male
Age: 72
Background:
Wants to return to sport activity. Underwent coronary CT angiography (CCTA) Has high blood cholesterol
and mild carotid plaque.
A long very severe coronary stenosis of the right coronary artery (RCA) from a partially calcified plaque was detected on mid RCA (CorEx software, Spimed-AI).
Subocclusive lesion was confirmed by invasive coronary angiography.
The patient was successfully treated by 2 stents on RCA.
This case illustrates that asymptomatic patients over 50 years of age may benefit from CCTA before returning
to sport activity.