**Anping Song \*, Zuoyu Wu, Xuehai Ding, Qian Hu and Xinyi Di**

School of Computer Engineering and Science, Shanghai University, Shanghai 200444, China;

zzzuo@i.shu.edu.cn (Z.W.); dinghai@shu.edu.cn (X.D.); ileven@shu.edu.cn (Q.H.); jsbluecat@shu.edu.cn (X.D.) **\*** Correspondence: apsong@shu.edu.cn; Tel.: +86-136-5167-5579; Fax: +86-6613-5550

Received: 27 October 2018; Accepted: 15 November 2018; Published: 16 November 2018

**Abstract:** Facial nerve paralysis (FNP) is the most common form of facial nerve damage, which leads to significant physical pain and abnormal function in patients. Traditional FNP detection methods are based on visual diagnosis, which relies solely on the physician's assessment. The use of objective measurements can reduce the frequency of errors which are caused by subjective methods. Hence, a fast, accurate, and objective computer method for FNP classification is proposed that uses a single Convolutional neural network (CNN), trained end-to-end directly from images, with only pixels and disease labels as inputs. We trained the CNN using a dataset of 1049 clinical images and divided the dataset into 7 categories based on classification standards with the help of neurologists. We tested its performance against the neurologists' ground truth, and our results matched the neurologists' level with 97.5% accuracy.

**Keywords:** facial image analysis; facial nerve paralysis; deep convolutional neural networks; image classification
