Figure 1 | Scientific Reports

Figure 1

From: Automatic diagnosis of retention pseudocyst in the maxillary sinus on panoramic radiographs using a convolutional neural network algorithm

Figure 1

Model architecture and annotation. Overall architecture of the proposed model from EfficientDet-D4 (a). Conv, convolutional network; MBConv, mobile inverted bottleneck convolution block; BIFPN, bi-feature pyramid network. Examples of annotated panoramic radiographs and annotation information (b). The blue boxes indicate class 0 (healthy), the red box indicates class 1 (retention pseudocyst), and the yellow box indicates class 2 (cyst or tumor). The information was composed of the location, represented as the upper left (X1, Y1) and the lower right (X2, Y2), and each class name.

Back to article page