clinical study

The utility of deep learning: evaluation of a convolutional neural network for detection of intracranial bleeds on non-contrast head computed tomography studies

Materials & Methods

The algorithm was tested on 7,112 non-contrast head CTs acquired during 2016–2017 from two, large urban academic and trauma centers. 


Promising results of a scalable and clinically pragmatic AI model tested on a large set of real-world data from high-volume medical centers. Results showed a 98% level of accuracy, with 95% sensitivity and over 98% specificity. 


The AI algorithm had high levels of accuracy (98%), sensitivity (95%) and over 98% specificity (98%)