top of page
Degradé site web.png

ReduCe unCerTAinty

One heart exam and that's it !

Spimed-AI has developed an innovative artificial intelligence application, called CorEx.

CorEx Automatically analyzes Coronary CT images (CCTA), independently detecting and classifying coronary lesions quantitatively as an expert second reader.

CorEx prioritizes and draws your attention to at-risk patients while expediting the processing of normal exams with increased confidence.

Thanks to CorEx, numerous invasive and costly examinations can be replaced by

a single, comprehensive exam.

Écran d'ordinateur affichant l'interface CorEx

2 modules, 1 exam     1 relevant result 


The solution automatically classifies stenosis according to the international CAD-RADS classification. With CorEx, you can see in seconds which artery is damaged and the severity of the obstruction.

For research use only


An attenuation of more than 80% of the blood flow through the stenosis is the common threshold to consider the stenosis as significant and dangerous. Beyond this point, surgery is recommended. The most common procedures are the stent and the bypass.

For each coronary, the solution predicts if this important threshold is reached or not.

Degradé site web.png


Accuracy on CorEx

Validated by a study

30 000

Database to  

train our AI


Hospitals and clinics 

Using our solution

CorEx Benefits

An international vision

CorEx is used in over 30 centers worldwide.

Already present in Europe and in the USA (in research only), we continue to conquer the different continents.

CE marked since 2021, we are waiting for the FDA approval in 2023. 

carte 2 bleu.png
Carte du monde en pointillé indiquant les centres de test de CorEx (USA, Europe, Japon)


Pr Jean-François DEUX

Pr Jean-François DEUX, Radiologist

"I think this system could bring more to daily interpretation cardiac scans, at the beginning radiologist by confirming his analysis, but also with the more experienced radiologist by prioritizing, for example, the reading of examinations via an automated analysis prior to the human radiological reading."

bottom of page