Figure 5 | Scientific Reports

Figure 5

From: A joint convolutional-recurrent neural network with an attention mechanism for detecting intracranial hemorrhage on noncontrast head CT

Figure 5

The presentative images of different patients in whom the model predictions were wrong (The images were created by the authors using open-source software, Matplotlib v3.5, Python v3). The original (the upper left) and corresponding normgrad images (the upper right) with a false-positive prediction are shown. In addition, the model overlooked the minor subarachnoid hemorrhage in the left frontal lobe (the lower left); the model missed the minor subarachnoid hemorrhage in the frontal lobe and subdural hemorrhage in the frontotemporal area (the lower right).

Back to article page