StrokeNeXt: A Siamese-encoder Approach for Brain Stroke Classification in Computed Tomography Imagery

2026-02-16

AI Summary

This work addresses the automatic detection of stroke and its subtype classification (ischemic versus hemorrhagic) in computed tomography (CT) images by proposing a lightweight and efficient Siamese-ConvNeXt architecture. The model employs dual-branch ConvNeXt encoders to extract multi-scale features, which are then fused through a 1D convolutional decoder, a bottleneck projection module, and a compact classification head. Evaluated on a dataset of 6,774 CT images, the method achieves an accuracy and F1 score of 0.988, significantly outperforming both CNN- and Transformer-based baselines. It also demonstrates low inference latency, rapid convergence, high sensitivity and specificity across classes, and excellent calibration of predictive confidence, making it well-suited for clinical deployment.

Technology Category

• Computer Vision: Object Detection & Categorization
• Machine Learning: Deep Neural Architectures and Foundation Models
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

We present StrokeNeXt, a model for stroke classification in 2D Computed Tomography (CT) images. StrokeNeXt employs a dual-branch design with two ConvNeXt encoders, whose features are fused through a lightweight convolutional decoder based on stacked 1D operations, including a bottleneck projection and transformation layers, and a compact classification head. The model is evaluated on a curated dataset of 6,774 CT images, addressing both stroke detection and subtype classification between ischemic and hemorrhage cases. StrokeNeXt consistently outperforms convolutional and Transformer-based baselines, reaching accuracies and F1-scores of up to 0.988. Paired statistical tests confirm that the performance gains are statistically significant, while class-wise sensitivity and specificity demonstrate robust behavior across diagnostic categories. Calibration analysis shows reduced prediction error compared to competing methods, and confusion matrix results indicate low misclassification rates. In addition, the model exhibits low inference time and fast convergence.

Problem

• stroke classification
• computed tomography
• ischemic stroke
• hemorrhagic stroke
• medical image analysis

Innovation

• Siamese-encoder
• ConvNeXt
• lightweight decoder
• stroke classification
• CT imagery