Figure 4 | Scientific Reports

Figure 4

From: A pilot trial of Convolution Neural Network for automatic retention-monitoring of capsule endoscopes in the stomach and duodenal bulb

Figure 4

The neural network model. The endoscopic image classification network is a modified Convolutional Neural Network based on LeNet. The endoscopic image is sent directly to the network as an input, and the neural network can directly calculate the probability of the stomach and the small intestines. According to the probability score, it can be discriminated whether the input image is the stomach/duodenal bulb or the DSD.

Back to article page