Quadratic Convolution-based YOLOv8 (Q-YOLOv8) for localization of intracranial hemorrhage from head CT images

Abdesselam Ferdi, Said Benierbah, Amir Nakib, Youcef Ferdi, Abdelmalik Taleb-Ahmed

Abstract: Highlights•ICH is a fatal disease that requires early and rapid diagnosis by using CT scans.•Accurate ICH localization is challenging, therefore the Q-YOLOv8 detector is proposed.•A novel YOLO algorithm based on the quadratic convolution and YOLOv8 detector is developed to improve the detection performance of existing ICH detectors.•The detector provided important information about the ICH by predicting a bounding box around the infected regions as well as the hemorrhage type and confidence score.