Amenorrhea, the absence or abnormal cessation of menstruation, is a condition that affects many women worldwide. Medicinal plants have been traditionally used for their potential therapeutic properties in treating various health conditions, including amenorrhea. There is a need to identify potential medicinal plants that could contribute to the development of new treatments that are safe, effective, and accessible to women who suffer from this condition. Natural amenorrhea treatments can indeed be seen as a safer and more accessible option for some individuals. These treatments typically involve the use of medicinal plants, herbs, or natural remedies that are believed to have beneficial effects on the female reproductive system. In this study, we propose a novel approach for the identification of medicinal plants specifically targeted for amenorrhea using a deep fused neural network ensemble with bagged trees, which has achieved remarkable accuracy. The DenseNet201-based ensemble model showcases an outstanding accuracy rate of 100%, underscoring its potential as a reliable and efficient tool for the selection and utilization of medicinal plants to address amenorrhea. Future improvements and practical deployment of this model hold considerable promise in enhancing the overall healthcare and well-being of women affected by amenorrhea. The resulting model can be deployed to predict the properties of new, unseen medicinal plants in the form of a mobile app. The mobile app allows users to input relevant plant characteristics or take pictures of plants for instant identification.