Intracranial Hemorrhage Detection

Intracranial hemorrhage is an urgent condition detectable from non-contrast head CT exams. We have developed a deep learning algorithm that can detect ICH on non-contrast head CT images in just a few seconds. The algorithm can be incorporated into a software medical device to help third party worklist systems triage and prioritize reading of images for patients with ICH, thus improving the likelihood of prompt detection and intervention. 

This project detects ICH by analyzing non-contrast CT images. Integrated with a third-party worklist application, the algorithm output of whether a study suggests ICH condition is sent to the worklist application via the software’s application program interface. The worklist application can use this information for prioritization purposes. 

  • ICH detection
  • Provides hemorrhage location and sizing
  • Classifies five subtypes of hemorrhage
  • Designed for worklist prioritization and triage
  • Highly interoperable
Precise Diagnosis of Intracranial Hemorrhage and Subtypes Using a Three-dimensional Joint Convolutional and Recurrent Neural Network Ye H*, Gao F* (*Equal contribution), Yin Y, Guo D, Zhao P, et al., European Radiology, March, 2019
Return to Portfolio