Doctorate Muhammad Ahmad: Student at University of Electronic Science and Technology of China, China

Muhammad Ahmad is a passionate AI researcher and software engineer with expertise in generative AI, deep learning, and medical image analysis. Currently, he is pursuing a Master’s degree in Information and Communication Engineering at the University of Electronic Science and Technology of China (UESTC), where he maintains a strong academic record with a GPA of 3.54/4.0. Muhammad combines his solid theoretical background with hands-on experience in developing innovative AI models, focusing on healthcare applications to improve diagnostic accuracy and patient outcomes.

Online Profiles

ORCID Profile

Education

Muhammad completed his Bachelor of Science in Computer Science from the University of South Asia in Lahore, Pakistan, graduating with a CGPA of 3.16/4.0. During his undergraduate studies, he conducted a notable final year project on Walmart Weekly Sales Prediction using machine learning techniques. Currently, he is advancing his knowledge and research capabilities through a fully funded Master’s program at UESTC, China, where his studies focus on cutting-edge topics such as large language models, generative AI, and medical imaging technologies.

Research Focus

His research is concentrated on the application of deep learning and generative AI methods to medical imaging, particularly focusing on neurodegenerative diseases and brain tumor diagnostics. Muhammad explores the fusion of local and global image features using hybrid architectures and leverages large language models (LLMs) to improve the interpretability and accuracy of medical image analysis, aiming to develop scalable AI solutions for early disease detection and prognosis.

Experience

Muhammad has garnered practical industry experience as a Software Engineer specializing in Artificial Intelligence at E-teleQuote Inc. in Florida, USA, where he led projects on generative AI, real-time chatbot development, and speech-to-text and sentiment analysis technologies. Prior to this, he completed a Machine Learning internship at Quid Sol in Lahore, Pakistan, where he built data pipelines, implemented machine learning algorithms for object detection and tracking, and developed custom deep learning models, strengthening his skills in feature engineering and AI optimization.

Research Timeline

In June 2025, Muhammad published a peer-reviewed article on a hybrid deep learning architecture with adaptive feature fusion for Alzheimer’s disease classification in Brain Sciences. In May 2025, he submitted a manuscript detailing a novel method for dynamic fusion of local and global features aimed at improving brain tumor diagnosis to the International Journal of Machine Learning and Cybernetics. These works reflect his commitment to advancing AI-driven medical research through innovative model designs and rigorous evaluation.

Awards & Honors

Muhammad’s academic excellence and community contributions have been recognized through several awards, including a fully funded scholarship for his Master’s degree at UESTC. He earned a semester scholarship for excellence in a machine learning workshop and received volunteer recognition for his active role in community welfare initiatives through the Rooh Foundation. Additionally, he has demonstrated his technical prowess by winning first and second positions in competitive web development contests hosted by COMSATS University and Superior University, respectively.

Top-Noted Publication

Muhammad’s most notable publication, “Hybrid Deep Learning Architecture with Adaptive Feature Fusion for Multi-Stage Alzheimer’s Disease Classification,” published in Brain Sciences in 2025, presents an innovative approach that integrates multiple feature sets for more accurate staging of Alzheimer’s disease. This research contributes significantly to the field of AI-assisted medical diagnostics by enhancing feature representation and classification performance, offering a promising tool for clinical applications.

Hybrid Deep Learning Architecture with Adaptive Feature Fusion for Multi-Stage Alzheimer’s Disease Classification
Brain Sciences | 2025-06-06
DOI: 10.3390/brainsci15060612
Contributors: Ahmad Muhammad, Qi Jin, Osman Elwasila, Yonis Gulzar

Abstract (expanded):
This article introduces a novel hybrid deep learning architecture designed to enhance the classification of Alzheimer’s disease (AD) through the integration of adaptive feature fusion. The model combines both global and local features from neuroimaging data, optimizing classification performance across multiple stages of Alzheimer’s disease. The study leverages convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to process and fuse brain images for a more accurate disease diagnosis and staging. By enhancing feature extraction with adaptive fusion techniques, the architecture outperforms traditional methods in terms of accuracy and reliability.

Key Contributions:

1. Hybrid Model Design: The integration of CNNs for feature extraction and RNNs for sequence learning provides a comprehensive framework for Alzheimer’s classification.

2. Adaptive Feature Fusion: The study proposes a dynamic fusion strategy that adapts to various stages of the disease, improving classification precision.

3. Stage-Specific Diagnosis: The model successfully categorizes Alzheimer’s at different stages, addressing the clinical challenge of early-stage detection.

Impact & Relevance:
This work is a significant step forward in the use of AI for medical diagnostics, particularly for neurodegenerative diseases. The proposed model has the potential to aid clinicians in diagnosing Alzheimer’s disease at earlier stages, leading to better patient outcomes. It highlights the growing role of deep learning in medical image analysis and opens doors for future research on hybrid architectures in healthcare.