DDTNet: A Dense Network with Dendritic Learning for Brain Tumor Classification

The rapid advancement of artificial intelligence has transformed the field of medical image analysis, particularly in the detection of brain tumors. The implementation of deep learning techniques has significantly enhanced the precision of tumor localization and size estimation, thereby improving the accuracy of tumor identification and pre-diagnosis. However, existing methods often suffer from poor edge detection, leading to the poor classification performance due to weak nonlinear feature representations. To alleviate this problem, we propose a novel architecture termed Dendritic Dense Tumor-detection Network (DDTNet). This architecture introduces dendritic neuron module to enhance nonlinear feature representation at the end of dense networks. By leveraging the superior non-linear capabilities of DNMs to emulate dendritic neuron behavior, DDTNet facilitates improved feature classification. Experimental results on publicly available datasets demonstrate that DDTNet outperforms both traditional DenseNet and other state-of-the-art methods in terms of accuracy and computational efficiency. For the brain tumor detection classification problem, our model achieves an accuracy of 96.3 % and an F1 score of 96.2 %. This study suggests that utilizing DNM as classifiers could facilitate the development of more effective deep learning models for classification tasks.