Hassan Ali

Hassan Ali

Auto-generated profile, based on publicly available information

Hassan Ali

College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics (NUAA), Nanjing, China

Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
16 March 2026 in IEEE Access
Verified
Journal Article
Open access
HAHassan Ali
Hassan Ali
Abid MehmoodAbid Mehmood
Abid Mehmood
DTDavid Noule Tolno
David Noule Tolno
SSSery Gahouidi Junior Thierry S
Sery Gahouidi Junior Thierry S
MSMuhammad Saeed
Muhammad Saeed
NANaeem Ahmed
Naeem Ahmed
Accurate segmentation of cardiac structures in 2D echocardiography is essential for diagnosing cardiovascular disease and computing clinical metrics such as chamber volumes and ejection fraction. Conventional U-Net architectures excel at extracting local spatial features but struggle with long-range dependencies inherent in noisy ultrasound images, while pure Transformer-based models capture global context at the expense of fine boundary detail. To address these limitations, we propose KMS-Net, a novel hybrid segmentation architecture that integrates Kolmogorov–Arnold Networks (KANs), a class of learnable, spline-based function approximators that replace fixed activation functions with trainable nonlinear mappings, alongside multi-scale attention mechanisms. Specifically, spline-based KAN layers (grid size of 7) provide highly expressive nonlinear feature modeling. Residual convolutional blocks incorporate Efficient Multi-Scale Attention (EMA) and a KAN-based Atrous Spatial Pyramid Pooling with Squeeze-and-Excitation module (KASPPS) that aggregates rich multi-scale contextual information. Selective State-Space (SS2D) layers enable computationally efficient long-range spatial dependency modeling, and Multi-Scale Attention Gates (MSAG) refine skip connections for precise boundary delineation. Evaluated on the CAMUS dataset (500 patients, two-chamber/four-chamber views at end-diastole/end-systole), KMS-Net achieves mean Dice scores of 0.9417 and 0.9241 (2-CH view) and 0.9518 and 0.9328 (4-CH view), outperforming U-Net variants, TransUNet, Swin-UNet, U-Mamba, and recent KAN-based models. The proposed method further attains state-of-the-art performance on the ACDC cardiac MRI benchmark (mean Dice: 92.65%, HD95: 1.08 mm) and demonstrates strong zero-shot generalization to the EchoNet-Dynamic dataset, confirming its robustness across diverse imaging modalities and clinical settings.

1 June 2026 in IEEE Access
Verified
Journal Article
Open access
HAHassan Ali
Hassan Ali
Muhammad Asghar KhanMuhammad Asghar Khan
Muhammad Asghar Khan
GBGordana Barb
Gordana Barb
ASAndrei Marius Silaghi
Andrei Marius Silaghi
Breast cancer histopathological classification remains a challenging task due to small patch resolution, staining inconsistencies, magnification variability, and class imbalance in datasets such as BreakHis and IDC, which hinder accurate malignancy detection. This paper proposes AttFKAN-DEQ, a novel framework that integrates Fourier Kolmogorov-Arnold Networks (FKANs) with Deep Equilibrium Models (DEQs) for binary benign-versus-malignant classification. Unlike spline-based formulations, FKANs model univariate edge functions using finite trigonometric series with trainable sine and cosine coefficients, enabling robust multi-scale spectral representations of heterogeneous tissue morphology. To enhance feature discrimination, a Lightweight Convolutional Block Attention Module (LCBAM) is incorporated within the FKAN residual block to perform channel-wise and spatial feature recalibration during each DEQ iteration using depthwise separable convolutions. This mechanism emphasizes diagnostically relevant regions while suppressing irrelevant background variations. Furthermore, the DEQ mechanism implicitly models infinite network depth through fixed-point iterations at constant memory cost, allowing stable and progressive latent feature refinement. The proposed framework was evaluated using five-fold stratified cross-validation with standard data augmentation. AttFKAN-DEQ achieved 96.60% accuracy and 98.44% AUC on the BreakHis dataset and 94.52% accuracy with 97.31% AUC on the IDC dataset. Experimental results demonstrate that the proposed model outperforms DEQ-KAN, ResNet variants, DenseNet-121, EfficientNet-B0, Vision Transformer, Swin Transformer, ConvNeXt, and MobileViT across multiple evaluation metrics. Ablation studies further confirm the individual and combined contributions of Fourier parameterization, lightweight attention, and equilibrium-based feature refinement to the overall performance. Despite containing only 6.0M parameters and requiring 0.60 GFLOPs, AttFKAN-DEQ achieves high diagnostic accuracy with strong computational efficiency, highlighting its potential for deployment in real-world clinical histopathology applications.

1 January 2026 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Open access
HAHassan Ali
Hassan Ali
Muhammad Asghar KhanMuhammad Asghar Khan
Muhammad Asghar Khan
GBGordana Barb
Gordana Barb
ASAndrei Marius Silaghi
Andrei Marius Silaghi

26 September 2026 in Biomedical Signal Processing and Control
Verified
Journal Article
Open access
HAHassan Ali
Hassan Ali
Muhammad Asghar KhanMuhammad Asghar Khan
Muhammad Asghar Khan
GBGordana Barb
Gordana Barb
ASAndrei Marius Silaghi
Andrei Marius Silaghi