Teleradiology services provider Virtual Radiologic (vRad) said that its deep-learning artificial intelligence (AI) algorithm has been used to identify potential intracranial hemorrhaging on CT images in real-time.
The firm's physicians used anonymized data from vRad's clinical database to train the AI software to search, analyze, and correctly recall cases of intracranial hemorrhaging, vRad said. When the algorithm is incorporated into the firm's telemedicine workflow, it will "flag" all potential intracranial hemorrhaging cases and automatically prioritize the study within the radiologist's reading queue. Cases with potential intracranial hemorrhaging can be assigned to the most appropriately trained/experienced radiologist for quick diagnosis and communication of findings to the attending physician, vRad said.
The company said it would begin the process to obtain regulatory approval for the algorithm before the end of the year. Once the algorithm is approved, vRad intends to implement this process into its telemedicine workflow. More than 5,000 studies could be identified by the algorithm in 2016, according to the firm.
