Q-VISION is a three-year research program developing a real-time AI framework to predict and mitigate post-operative complications in robot-assisted nephrectomy for kidney cancer. The project is funded by the Qatar Research, Development and Innovation (QRDI) Council under grant ARG01-0522-230266, with Hamad Medical Corporation (HMC) as the lead institution, in collaboration with Qatar University (QU) and Hamad Bin Khalifa University (HBKU).

Our research integrates scene segmentation, surgical vision-language reasoning, and temporal workflow analysis into a unified clinician-facing platform, with two translational assets currently in active development: ‘SurgXpert’, a browser-based inference platform, and ‘SurgPixel Studio’, a boundary-aware annotation tool powered by our CVPR 2026 main-track work.

Outputs so far are based on publicly available datasets, and include multiple peer-reviewed publications spanning CVPR, MIDL, ECAI, CBMS, and CASE (including Qatar-based first-authorships at flagship international venues). The overall aim is development of a complication-prediction prototype aligned to HMC’s clinical deployment pathways.

Figure: Real-time surgical AI framework for kidney cancer surgery.

With the emergence of 5G networks and the anticipation of 6G networks, the communication performance requirements are becoming increasingly stringent, and the networks are approaching Shannon's theoretical capacity limits. Therefore, a new communication paradigm has emerged, namely, semantic communication. Although the term itself is not new, research in this field has increased recently due to the need of sending huge amounts of data with limited communication resources and due to the emergence of Artificial Intelligence and the increasing computational capabilities of the communicating devices.

In the past few years, research in the field of semantic communication systems has tackled the different approaches of semantic communication, different types of data and how they can be transmitted semantically, and also the challenges and limitations of semantic communication systems were highlighted in many works. However, semantic communication research is still in its early stages and it is still not clear whether we should adopt this paradigm when designing a new communication system. Additionally, the reliability of semantic communication systems is still not thoroughly investigated and compared with traditional communication systems. Finally, most of the currently proposed semantic communication systems are based on machine learning models that are trained on specific datasets; this does not guarantee their generalization capability.

Therefore, the main objective of this project is to investigate techniques to enhance the reliability and generalization of semantic communication techniques and to compare the performance of networks with and without the use of semantic communication systems.

Figure: Reliable semantic communication framework for 6G and beyond networks.

This project aims to develop an advanced AI-driven peptide engineering and validation platform to target anti-apoptotic BCL-2 family proteins in colorectal cancer (CRC). Colorectal cancer remains a major global health challenge, largely due to resistance to therapy caused by dysregulated apoptosis. Overexpression of BCL-2 proteins allows cancer cells to evade cell death, highlighting the need for innovative therapeutic strategies.

To address this, the proposed research integrates artificial intelligence, computational modeling, and experimental validation to design novel pro-apoptotic peptides capable of restoring programmed cell death in cancer cells. The platform will utilize cutting-edge AI approaches, including generative modeling and inverse protein folding, followed by molecular docking, molecular dynamics simulations, and binding free energy analysis to identify high-affinity peptide candidates.

Selected peptides will be further optimized and experimentally validated using colorectal cancer cell models through assays such as MTT, flow cytometry, Western blotting, and biophysical techniques. Beyond developing novel therapeutic candidates, the project aims to establish a scalable and reproducible AI-driven drug discovery framework in Qatar, contributing to national biopharmaceutical innovation.

This research project aims to explore the potential of mixed reality (MR) and artificial intelligence in supporting clinical dentistry through visually guided workflows. The focus is on integrating real time imaging, AI based segmentation, and spatial computing to enable digital information to be overlaid onto the clinical field during treatment.

The project will investigate the development of models capable of identifying teeth and anatomical structures from intra oral video in near real time. These outputs will be incorporated into an MR environment, with the intention of providing visual references that may assist clinicians during procedures such as tooth preparation, implant placement, and endodontic access.

In parallel, the research will address practical considerations related to clinical implementation, including data quality, alignment accuracy, and system stability under routine working conditions. Efforts will also be made to define appropriate validation approaches to assess performance and reliability. Overall, the project seeks to better understand how digital planning and real time clinical execution can be more closely connected.

This project develops an AI-powered drone system to support search and rescue operations in disaster situations such as earthquakes or building collapses. The system uses multiple intelligent drones equipped with advanced sensing technologies to detect signs of human presence even under rubble or in difficult environments.

Artificial intelligence enables the drones to autonomously explore damaged areas, coordinate with each other, and identify locations where survivors may be trapped. By combining AI, sensing, and autonomous navigation, this research aims to assist emergency teams in locating survivors faster, improving rescue efficiency, and ultimately saving more lives during critical situations.

Figure: AI-powered autonomous drone coordination for search and rescue.

In a world where technology is transforming how we learn, assess, and evolve, the ability to measure Arabic writing skills remains a critical challenge. In this project, funded by QRDI and supported by the Ministry of Education, we aim to develop the first AI-powered system to automatically score Arabic essays of high school and university students over different writing traits (e.g., organization and development of ideas).

Since the project started in 2023, we have constructed a unique Arabic annotated essay dataset, LAILA, comprising about 8,000 student essays from 24 Qatari schools. We also proposed two novel essay scoring frameworks, TRATES and MAPLE, that achieve state-of-the-art performance. Furthermore, we have developed an online scoring platform, Qayyem, that can serve teachers and instructors in assessing Arabic writing proficiency.

The research has been published in top-tier conferences such as ACM SIGIR, ACL, and EACL. The project is led by Prof. Tamer Elsayed (QU), Dr. Houda Bouamor (CMUQ), and Dr. Walid Massoud (QU). For further information, visit qayyem.qu.edu.qa.

Federated learning (FL) offers a promising approach to collaborative machine learning by enabling participants to train models locally while sharing only updates instead of raw data, thereby enhancing privacy. However, FL systems remain highly vulnerable to adversarial threats at both the client and server levels. At the node level, malicious actors can launch data and model poisoning attacks. In contrast, at the server level, even semi-honest aggregation servers may exploit model updates to conduct inference attacks and extract sensitive information. Existing solutions predominantly address one of these vulnerabilities in isolation, leaving critical security gaps in practical deployments.This project proposes the design and implementation of a comprehensive security framework that simultaneously mitigates both client-side and server-side threats in FL systems. The project will combine secure multi-party computation (MPC) to safeguard the integrity and confidentiality of model updates during aggregation, with differential privacy (DP) to limit information leakage and counter inference and poisoning attacks. The project will validate the framework through deployment in an FL environment using appliance-level micro-moment energy consumption data, thereby demonstrating its feasibility, robustness, and practical impact. Ultimately, this work seeks to advance secure and trustworthy federated learning systems by addressing multi-level adversarial risks in a unified manner.

Figure: PRIVMAC Federated Learning Framework.

This innovation project aims to leverage Artificial Intelligence (AI) and Large Language Models (LLMs) to enable future intelligent robotics applications and autonomous systems. The project focuses on combining AI reasoning, computer vision, sensor fusion, edge computing, and robotic control systems to create adaptive, collaborative, and decision-capable robots.

By enabling robots to understand natural language instructions, interpret complex environments, and autonomously plan actions, the platform will accelerate the development of next-generation robotics solutions with enhanced autonomy, flexibility, safety, and operational efficiency for industrial, healthcare, logistics, and service applications.

Figure: LLM-powered robotics platform for autonomous industrial applications.

Dairy sector has grown remarkably in recent years, yet the environmental cost of producing milk, yoghurt, cheese, and other products, from greenhouse gas emissions to water consumption and land use, remains poorly understood at the local level. This project sets out to change that.

Led by researchers at Qatar University in collaboration multiple diary producers, the project aims to build a first-of-its-kind AI-powered environmental assessment tool tailored specifically to Qatar's dairy supply chains. At its core is a novel hybrid Life Cycle Assessment (LCA) framework that combines two established environmental analysis methods to capture impacts across the full production chain from farm inputs all the way to the finished product on the shelf.

What makes this approach distinctive is the integration of artificial intelligence. By applying machine learning and deep learning algorithms, the team will dramatically improve the accuracy and speed of environmental impact predictions across a wide range of production scenarios. The result will be a user-friendly, interactive software tool that allows policymakers, industry stakeholders, and sustainability planners to model different decisions and instantly understand their environmental consequences.

Beyond the technology itself, the project will produce concrete, evidence-based recommendations for reducing the dairy industry's environmental footprint, informing future policy, guiding industry practices, and supporting Qatar's broader national goals for food security and sustainable development.

Figure: AI-Powered Sustainability Assessment for Dairy Production.

This project focuses on strengthening the capacity of humanitarian organizations in Qatar to use artificial intelligence (AI) in a responsible, practical, and impactful way. It responds to the growing need for evidence-based and context-specific approaches to AI adoption in high-stakes humanitarian environments.

The project generated a comprehensive evidence base through two large scoping reviews, mapping both the ethical principles and real-world applications of AI in humanitarian settings. Findings highlighted a clear gap between high-level ethical guidance and its practical implementation, as well as uneven adoption of AI tools across different contexts.

To understand readiness in Qatar, the project conducted a national survey and in-depth interviews with humanitarian stakeholders. Results showed strong interest in AI, particularly for data management, needs assessment, and monitoring, but also revealed key challenges, including limited technical capacity, governance gaps, and concerns around data privacy and trust.

A key output of the project is the development of practical ethical guidance, designed to support organizations in implementing AI responsibly across real-world humanitarian operations. This guidance helps bridge the gap between global frameworks and on-the-ground application.

The project also established a national Technical Working Group (TWG) chaired by the Ministry of Foreign Affairs and co-led by Qatar Funds for Development, bringing together stakeholders from government, academia, and international organizations to co-design solutions and guide AI adoption in alignment with national priorities.

In addition, the project led to the creation of the Humanitarian AI Observatory, an online platform that curates policies, case studies, and tools to support knowledge-sharing and capacity building. The Observatory serves as a central hub for advancing evidence-based and ethical AI use in the region.

Overall, the project provides a foundation for responsible AI adoption by translating research into practical guidance, identifying priority use cases, and contributing to the development of national strategies and capacity-building initiatives. It positions Qatar as an emerging leader in advancing ethical and evidence-based AI in humanitarian action.

Figure: Strengthening AI in the Humanitarian Sector in Qatar.

The transition toward 6G wireless networks demands new approaches for efficient data transmission. As current systems approach theoretical limits for raw data compression, transmitting actionable knowledge instead of traditional bit level data becomes essential. Legacy communication frameworks based on Shannon separation principles face severe challenges with high dimensional machine observations. Increasing compression discards critical semantic features and pushes error correcting codes to their limits, causing communication to collapse in what is known as the digital cliff effect.

This project addresses this bandwidth limitation by shifting the objective from pixel level reconstruction to task relevant knowledge preservation. By leveraging Deep Joint Source Channel Coding (Deep JSCC), the research demonstrates how autoencoders can jointly optimize compression and channel robustness in a single end to end process. This approach leads to resilient knowledge transmission that enables autonomous agents to maintain high performance even as channel conditions degrade compared to traditional techniques.

The project extends the evaluation of Deep JSCC by characterizing bottleneck configurations under 6G Rayleigh fading channels, specifically benchmarking convolutional neural networks (CNN) and Vision Transformer (ViT) architectures against traditional JPEG source and Low-Density Parity Check (LDPC) channel encoders. By measuring reward preservation for remote reinforcement learning agents across these varied conditions, this study identifies the most effective architectures for ensuring reliable goal-oriented communication.

Figure: Goal-oriented semantic communication with Deep JSCC.

Anomalous energy consumption drains billions from utilities worldwide. This project explores whether modern large language models (LLMs) can produce human-readable explanations for energy theft and anomalous behavior, rather than acting as "black box" detectors.

Using three years of smart-meter readings, the research benchmarks vision LLMs like Gemini and Qwen against classical ML baselines. The goal is to demonstrate that lightweight open-source LLMs can match proprietary models in accuracy while providing explanations that help field crews identify exact suspicious hours and attack patterns.

Figure: Explainable AI framework for detecting energy theft in smart grids.

The rapid emergence of extended reality (XR) applications within the metaverse imposes stringent requirements on future wireless networks, including ultra-low latency, high data rates, and reliable quality of experience (QoE). At the same time, ensuring secure transmission of immersive XR data introduces additional overhead that can negatively impact network performance.

This project proposes a cross-layer optimization framework for secure XR data transmission in 6G-oriented metaverse environments, jointly considering communication efficiency, QoE, and security requirements. The proposed approach integrates adaptive bitrate control, dynamic compression strategies, and selective encryption mechanisms into a unified decision framework.

A reinforcement learning (RL)-based agent is developed to intelligently adjust transmission parameters based on real-time network conditions and user motion dynamics. The system aims to balance competing objectives, including latency, packet loss, visual quality, and security overhead, by learning optimal policies for resource allocation and protection levels. The framework is evaluated using realistic XR traffic datasets and simulated network conditions.

Figure: Secure XR pipeline with reinforcement learning for 6G metaverse.

The rapidly evolving landscape of wireless communication, particularly with the advent of 5G and future 6G networks, demands innovative approaches for efficient resource management. As networks become increasingly complex, the ability to optimize resource allocation in real-time while maintaining Quality of Service (QoS) across diverse service types becomes paramount.

This project focuses on addressing the issue of convergence acceleration in RL for wireless resource allocation. By leveraging expert knowledge transfer, we demonstrate how policies learned in specialized environments (such as eMBB, URLLC, and mMTC) can be reused by a learner agent to accelerate its convergence. This process leads to more efficient decision-making, reducing the time required to achieve optimal resource allocations.

Building on this foundation, the project extends the application of multi-agent reinforcement learning (MARL) by introducing a competency-based coordination model. This model dynamically adjusts the roles of agents based on their competencies, ensuring that each agent contributes in the most effective manner for its level of expertise.

Figure: Competency-based multi-agent RL for radio resource management.

This project focuses on developing a novel optical and AI-assisted platform to visualize and analyze nanobubbles in real-time. Nanobubbles are emerging as a promising technology for improving water and wastewater treatment, yet their behavior remains poorly understood due to limitations in existing measurement tools.

Our system integrates multiple laser wavelengths with advanced tracking algorithms to distinguish nanobubbles from other particles in complex environments. The collected data is enhanced using AI-based classification to improve accuracy. The project aims to optimize nanobubble-assisted processes such as dissolved air flotation (DAF), leading to more efficient and sustainable water treatment systems.

Figure: AI-assisted optical platform for real-time nanobubble tracking.

This innovation project aims to leverage Artificial Intelligence (AI) and Large Language Models (LLMs) to enable a new generation of intelligent Smart City applications and solutions. The project focuses on integrating AI, IoT, urban data platforms, and real-time analytics to improve city operations, mobility, sustainability, and public safety.

By combining multimodal AI reasoning with connected urban systems, the platform will support predictive decision-making, autonomous city operations, intelligent assistants, and data-driven governance to enhance quality of life, operational efficiency, and sustainable urban development in line with modern urban needs.

Figure: AI-driven urban data platform for predictive smart city governance.

This project develops a novel explainable and uncertainty-aware AI companion for pathologists that performs tumour grading while explicitly identifying uncertain cases for expert review. The first phase introduces CNTA (Cross-Modal Nucleus–Text Adapter), combining pathology foundation models with pathologist-informed textual descriptors.

Trained on over 150,000 oral squamous cell carcinoma nuclei, the framework achieves 88.4% balanced accuracy. Future phases will expand this methodology to thyroid FNAC diagnosis and introduce an interactive diagnostic chatbot to provide clinically grounded decision support for busy pathologists.

Figure: CNTA framework for confidence-aware histopathology diagnostic reasoning.

This project harnesses AI to build a population-specific predictor of autoimmune disease risk for the Qatari population. By integrating genomic and proteomic data from the Qatar Precision Health Initiative, the model overcomes the "blind spot" of European-centric risk tools.

Using XGBoost and SHAP explainable-AI tools, the model reveals the specific genetic variants and proteins driving risk in Qatari individuals. The outcome is an interactive predictive dashboard that translates complex multi-omics data into actionable clinical insights, advancing precision medicine in the region.

Qatar seeks to move healthcare beyond reactive treatment toward proactive, ICT-enabled prevention. This requires the development of autonomous E-health services over 5G/6G networks to support applications like remote surgery, survivor detection, elderly monitoring, and medical imaging.

To address challenges like large volumes of distributed medical data and strict low latency requirements, this project proposes a context-aware Health 4.0 platform. The platform integrates IoT, AI, NFV, edge/cloud computing, and collaborative learning to enable secure, efficient, and scalable healthcare services in line with Qatar National Vision 2030.

Figure: AI-driven edge platform for autonomous healthcare services.

Kidney stone disease is highly prevalent in Qatar. FaceStone is an AI-powered clinical interface that distills complex multi-omics molecular knowledge into a model that requires only routine clinical inputs.

By using knowledge distillation, the system eliminates the need for expensive omics profiling in primary care settings. FaceStone enables molecularly informed risk stratification for calcium oxalate kidney stones using data already collected in routine consultations, moving management from reactive diagnosis to proactive prevention.

Figure: FaceStone clinical interface for multi-omics informed kidney stone risk prediction.

Our project leverages the power of Artificial Intelligence and High-Performance Computing to unlock the complex biological mechanisms of the PTUPB drug. By integrating massive amounts of transcriptomic and proteomic data into advanced, GPU-optimized deep-learning models, we aim to decode intricate disease pathways.

This approach creates a powerful predictive engine designed to achieve two main goals: gaining a deeper understanding of PTUPB's role at a molecular level and rapidly identifying new therapeutic targets. Ultimately, this fusion of computational intelligence and biology will accelerate the drug discovery process and pave the way for cutting-edge precision medicine solutions.

Figure: AI-driven discovery pipeline for PTUPB therapeutic targets.

The purpose of this research project is to combine AI with Density Functional Theory (DFT) to facilitate the prediction of electronic, optical, and mechanical characteristics in emerging materials like perovskites and graphene nanoribbons.

By training machine learning algorithms on accurate DFT data (such as density of states and bandgap energy), this strategy eliminates computational constraints and ensures the efficient development of cutting-edge materials for future technology applications.

This study provides a controlled evaluation of modern RL algorithms for adaptive compression in latency-sensitive telesurgical communication systems. Telesurgery systems require stable, low-latency video transmission that remains synchronized with latency-critical haptic and control feedback. However, wireless bandwidth in realistic environments is inherently time-varying, making static compression strategies inadequate.

When video bitrate approaches or exceeds available channel capacity, delay can disrupt real-time synchronization. This work investigates reinforcement learning (RL) for adaptive video compression under stochastic wireless capacity conditions. A simulation framework models regime-based bandwidth variations, with delay computed as a nonlinear function of bitrate and instantaneous capacity.

Video quality is represented using a PSNR-based proxy that distinguishes compression-induced distortion from channel-related degradation. Two continuous-control reinforcement learning algorithms (Proximal Policy Optimization, Soft Actor-Critic) and one discrete-action algorithm (Deep Q-Network) are evaluated and compared in terms of convergence, delay–quality tradeoff, and stability under bandwidth fluctuations.

Figure: RL-based adaptive compression for telesurgery under stochastic bandwidth.

Stroke is the leading cause of disability worldwide. The HYDROTHERABOTS project addresses this by integrating Robotics, Artificial Intelligence, Augmented Reality, and Virtual Reality into a new generation of gait rehabilitation tools.

The team is developing two robotic walkers: a land-based VR-assisted system and an AR-assisted underwater walker designed for hydrotherapy. Both are equipped with sensors that feed AI models personalizing each patient's recovery. This approach aims to widen access to high-quality rehabilitation and give stroke patients a faster, more engaging path back to independent mobility.

Figure: HYDROTHERABOTS gait rehabilitation system integrating AI and VR/AR.

This project investigates large-scale foundation models for breast cancer diagnosis through joint modeling of heterogeneous clinical data, moving beyond single imaging modalities. We develop a multi-modal framework integrating mammography, ultrasound, and DCE-MRI with clinical metadata.

Leveraging pretrained vision–language foundation models as fine-tuned backbones on NVIDIA H100 GPUs, the system enables lesion detection, BI-RADS classification, and treatment-response forecasting. The project focuses on improving generalization, calibration, and clinical interpretability in low-data regimes.

Figure: Multi-modal foundation model framework for precision oncology.

Shock is a life-threatening condition of circulatory failure with mortality rates reaching up to 50%. Each hour of delayed intervention significantly increases mortality, yet routine ICU practice still relies on intermittent sampling that may miss rapid physiological changes.

This project develops a microneedle biosensor and a wristband vital-sign monitor coupled with an AI early detection system. Our team has trained machine learning models on the HiRID ICU dataset and is now validating them on a local Qatari ICU cohort from Hamad Medical Corporation. The goal is to provide a fine-tuned, prognostic AI model ready for clinical use in Qatar.

Figure: Real-time AI monitoring and early shock detection for ICU patients.

Antimicrobial resistance (AMR) has become a major global health threat. In Qatar, factors such as high population mobility and advanced tertiary healthcare systems facilitate the rapid spread of multidrug-resistant organisms. Addressing this challenge requires rapid, integrated, and scalable genomic surveillance systems.

An AI-enabled platform can address these gaps by leveraging advanced deep learning models to accurately predict AMR genes and enhance surveillance capabilities. This project integrates clinical, epidemiological, and genomic data to provide real-time insights and coordinated surveillance efforts, moving beyond fragmented bacterial whole genome sequencing (WGS) workflows.

Figure: AI-enabled genomic surveillance for antimicrobial resistance.

This project aims to evaluate Torosanin as a potential multi-target inhibitor for colorectal cancer (CRC) using an integrated computational approach. Given the complexity of CRC and the involvement of multiple dysregulated pathways, a multi-target strategy is essential for effective therapy.

The study will employ network pharmacology and bioinformatics analyses to identify Torosanin-associated targets and key CRC-related genes, followed by protein-protein interaction (PPI) network construction and GO/KEGG pathway enrichment to uncover underlying mechanisms. Subsequently, molecular docking will be performed to assess binding affinity with selected targets, and molecular dynamics (MD) simulations will evaluate the stability and dynamics of protein-ligand complexes.

Further analyses, including RMSD, RMSF, Rg, SASA, hydrogen bonding, and MM/GBSA or MM/PBSA binding free energy calculations, will be conducted to validate interactions. Additionally, physicochemical and ADMET predictions will assess drug-likeness and safety. Overall, this study will provide mechanistic insights into the multi-target anti-cancer potential of Torosanin and establish a computational foundation for future experimental validation and drug development.

The project proposes the development and evaluation of an AI-driven Virtual Clinic platform for interprofessional clinical training, titled Hakeem. It uses coordinated AI agents, including virtual patients, diagnostic and therapeutic reasoning agents, and team interaction monitors, to deliver dynamic, competency aligned interprofessional training tailored to Qatar’s healthcare context. The platform integrates large language model (LLM)-powered patient dialogue, AI-driven clinical decision pathways, and learning analytics to simulate complex patient scenarios and support effective team-based decision making for students from different healthcare professions such as medicine, pharmacy, nursing, nutrition, physiotherapy and others. The project combines technological innovation with rigorous educational evaluation. First, a systematic review and stakeholder co-design process will define technical and pedagogical requirements. Second, the platform will be developed by integrating adaptive conversational agents, multiagent scenario generation, realtime decision modeling, automated competency-based feedback, and teaminteraction monitoring capabilities. Third, its effectiveness will be evaluated through a randomized controlled trial at the Tamayuz Simulation Center at Qatar University comparing AI-driven training with traditional simulation-based interprofessional education. Finally, a qualitative study will examine trust, usability, and the integration of AI agents to inform the development of a national implementation roadmap.

By integrating intelligent agents with simulation-based training, the project aims to strengthen interprofessional collaboration skills, enhance clinical reasoning, and support the development of an AI-ready health workforce in Qatar.

Figure: Hakeem Virtual Clinic framework.

Endometrial cancer is the second most common gynecological malignancy, and accurate diagnosis is critical for appropriate treatment and improved outcomes. Conventional histopathological assessment can be subjective and variable. Deep learning offers a promising approach to enhance diagnostic precision. This project evaluates a deep learning model for predicting malignancy in endometrial cancer using digitized H&E-stained slides from Qatari patients. By leveraging locally generated data, the study aims to develop a region-specific model that reflects population-specific histopathological features, ultimately improving diagnostic accuracy and supporting personalized cancer care in Qatar.

We conducted an initial pilot machine learning (ML) analysis, with support from KINDI, using MobileNetV2 and Kaggle’s computational capabilities. MobileNetV2 model is developed my Google for embedded vision applications image classification, object detection, and semantic segmentation. This initial phase was done on 70 patients (1,312 H&E-stained histopathological slides).

Currently, the PhD student is working on increase number of patients and scanning more H&E slides. We plan to implement a robust ML pipeline, enabling more sophisticated modeling, validation, and performance optimization.