The field of image segmentation and classification is moving towards more innovative and effective approaches. Recent developments have focused on improving the robustness and feature expressiveness of convolutional neural networks (CNNs) and exploring new architectures and loss functions for image segmentation tasks. Notably, there is a growing interest in utilizing spatial relationships and anatomical details in volumetric data to enhance segmentation accuracy. Additionally, semi-supervised learning techniques are being explored to alleviate the burden of acquiring high-quality annotated data. Some noteworthy papers include: NIRMAL Pooling, which achieves consistent improvements in image classification tasks by dynamically adjusting pooling parameters and applying a non-linear activation function. SRMA-Mamba, which introduces a novel network designed to model spatial relationships within complex anatomical structures of MRI volumes, enabling efficient volumetric segmentation of pathological liver structures. Diversity-enhanced Collaborative Mamba, which proposes a framework for semi-supervised medical image segmentation that explores and utilizes diversity from data, network, and feature perspectives. SCRNet, which proposes a novel Feature Aggregation Module that processes input features to establish a strong connection between long-range dependencies and local contextual information.