Fig. 2From: A deep learning model based on concatenation approach to predict the time to extract a mandibular third molar toothOverview of the concatenated model consisting of CNN part and MLP part. The CNN part learns image features through Convolutional Neural Network. The MLP part learns patients' clinical data through fully connected layers. Outputs from the MLP part and the CNN part are concatenated. Finally, there is an output layer that infers the extraction time, which is a positive integer, through fully connected layers. CNN Convolutional Neural Network; MLP Multilayer PerceptronBack to article page