Advanced machine learning techniques, including convolutional neural networks, are widely used in the healthcare field. Despite their high efficiency, convolutional neural networks are primarily designed to analyze images and have limitations when processing non-image data such as gene expression data. For this reason, few studies apply convolutional neural networks to omics data. Gene expression data in particular play a central role in the development of precision medicine, which aims to provide medical care tailored to the individual needs of the patients. The few studies that have been conducted in this field focus mainly on cancer and other diseases that could benefit from such an approach, such as atopic dermatitis, are not the subject of investigation. Therefore, in this work we employed a convolutional neural network to transcriptomic data collected from the skin of atopic dermatitis patients. As far as we know, this is the first paper reporting the application of a convolutional neural network to gene expression data of atopic dermatitis. The data were converted into images using the DeepInsight method. Furthermore, we also conceived a convolutional neural network model to distinguish between lesions and non-lesions and to extract the most significant genes. With two independent datasets, we achieved accuracy values of 75% and 85.59%. These results are considerably high to support the hypothesis that the identified genes may be involved in the development of atopic dermatitis lesions.