Gasparini Sara, Campolo Maurizio, Ieracitano Cosimo, Mammone Nadia, Ferlazzo Edoardo, Sueri Chiara, Tripodi Giovanbattista Gaspare, Aguglia Umberto, Morabito Francesco Carlo
Department of Medical and Surgical Sciences, Magna Græcia University, 88100 Catanzaro, Italy.
Regional Epilepsy Centre, Bianchi-Melacrino-Morelli Hospital, 89124 Reggio Calabria, Italy.
Entropy (Basel). 2018 Jan 23;20(2):43. doi: 10.3390/e20020043.
The use of a deep neural network scheme is proposed to help clinicians solve a difficult diagnosis problem in neurology. The proposed multilayer architecture includes a feature engineering step (from time-frequency transformation), a double compressing stage trained by unsupervised learning, and a classification stage trained by supervised learning. After fine-tuning, the deep network is able to discriminate well the class of patients from controls with around 90% sensitivity and specificity. This deep model gives better classification performance than some other standard discriminative learning algorithms. As in clinical problems there is a need for explaining decisions, an effort has been carried out to qualitatively justify the classification results. The main novelty of this paper is indeed to give an entropic interpretation of how the deep scheme works and reach the final decision.
提出使用深度神经网络方案来帮助临床医生解决神经学中一个困难的诊断问题。所提出的多层架构包括一个特征工程步骤（来自时频变换）、一个通过无监督学习训练的双重压缩阶段以及一个通过监督学习训练的分类阶段。经过微调后，深度网络能够以约90%的灵敏度和特异性很好地区分患者类别与对照组。这个深度模型比其他一些标准判别学习算法具有更好的分类性能。由于在临床问题中需要解释决策，因此已努力从定性角度证明分类结果的合理性。本文的主要新颖之处确实在于对深度方案如何工作并达成最终决策给出熵的解释。
