Ahmad Al-Kabbany

Ahmad Al-Kabbany

The rapid integration of generative AI into scientific research has exposed a critical gap in academic disclosure practice. Existing frameworks for reporting AI contributions are uniformly output-oriented -- they document what AI produced, not how the research unfolded. As a result, researchers who wish to report their AI collaboration honestly lack the tools to do so: no current framework can distinguish between a researcher who originated a research direction and one who adopted a direction proposed by AI, or between a researcher who critically evaluated AI-generated alternatives and one who accepted AI output without independent assessment. This gap is not a matter of compliance detail; it is a failure to capture the cognitive dynamics that determine what kind of intellectual contribution a paper actually represents. We propose PAIRED -- Process-Anchored Interaction Reporting for AI-Enabled Discovery -- a dual-facing framework that addresses this gap through four design principles: process orientation, which takes the decision point rather than the research product as the fundamental unit of documentation; dual-facing output, which derives a structured publisher disclosure from a prospective author log without double work; decision-point granularity, which operates between session-level coarseness and message-level impracticality; and artifact-triggered logging, which provides an auditable rule against selective omission. We demonstrate PAIRED through worked examples, discuss its limitations openly, and propose a model-assisted adoption pathway that embeds the framework's logging discipline directly into AI research platforms.

Ahmad Al-Kabbany, Esraa Kassem

This research is primarily concerned with the critical problem of synthesizing a structured Retrieval-Augmented Generation (RAG) system for advanced AI applications in the domain of swimming. As the integration of Artificial Intelligence in sports science matures, its applications in swimming have become increasingly diverse, spanning from real-time technical coaching and talent scouting to comprehensive performance profiling and the dynamic personalization of training periodization. Within this landscape, RAG-based systems represent a pivotal advancement in Large Language Model (LLM) enhanced swimming analysis, as they allow for the grounding of generative outputs in authoritative domain knowledge, thereby ensuring the credibility of AI-generated advice, contextually and technically. Despite this potential, building robust RAG systems using only real-world aquatic data presents significant challenges, including ethical constraints regarding athlete biometrics, and the high cost of manual expert labeling. To address these barriers, we propose a novel generative framework that leverages a multimodal knowledge base gathered across four dimensions: physiological data, physiological literature, kinematic sensor data, and unstructured domain expertise. Our proposed framework utilizes a multi-agent LLM architecture to synthesize a high-fidelity dataset of 1,864 validated "Question-Context-Answer" triplets-drawn from 1,914 drafts evaluated against 12 physiological soundness rules. By providing a structured, synthetic ground truth, this work establishes a foundational benchmark for trustworthy AI in aquatics. The outcomes of this research promise to enhance the reliability of automated coaching and open a plethora of future directions in "Meta-Agent" development and athletic profiling, ultimately bridging the gap between raw data engineering and practical sports science application.

Youssef Abuzeid, Shimaa El-Bana, Ahmad Al-Kabbany

The clinical deployment of deep learning models in medical imaging is severely hindered by domain shift. This challenge, where a high-performing model fails catastrophically on external datasets, is a critical barrier to trustworthy AI. Addressing this requires moving beyond simple performance metrics toward deeper understanding, making Explainable AI (XAI) an essential diagnostic tool in medical image analysis. We present a rigorous, two-phase approach to diagnose the generalization failure of state-of-the-art State-Space Models (SSMs), specifically UMamaba, applied to cerebrovascular segmentation. We first established a quantifiable domain gap between our Source (RSNA CTA Aneurysm) and Target (TopCoW Circle of Willis CT) datasets, noting significant differences in Z-resolution and background noise. The model's Dice score subsequently plummeted from 0.8604 (Source) to 0.2902 (Target). In the second phase, which is our core contribution, we utilized Seg-XRes-CAM to diagnose the cause of this failure. We quantified the model's focus by measuring the overlap between its attention maps and the Ground Truth segmentations, and between its attention maps and its own Prediction Mask. Our analysis proves the model failed to generalize because its attention mechanism abandoned true anatomical features in the Target domain. Quantitative metrics confirm the model's focus shifted away from the Ground Truth vessels (IoU~0.101 at 0.3 threshold) while still aligning with its own wrong predictions (IoU~0.282 at 0.3 threshold). This demonstrates the model learned spurious correlations, confirming XAI is a powerful diagnostic tool for identifying dataset bias in emerging architectures.

Aly Sabri Abdalla, Ahmad Al-Kabbany, Ehab F. Badran et al.

Vehicle-to-everything (V2X) networks support a variety of safety, entertainment, and commercial applications. This is realized by applying the principles of the Internet of Vehicles (IoV) to facilitate connectivity among vehicles and between vehicles and roadside units (RSUs). Network congestion management is essential for IoVs and it represents a significant concern due to its impact on improving the efficiency of transportation systems and providing reliable communication among vehicles for the timely delivery of safety-critical packets. This paper introduces a framework for proactive congestion management for IoV networks. We generate congestion scenarios and a data set to predict the congestion using LSTM. We present the framework and the packet congestion dataset. Simulation results using SUMO with NS3 demonstrate the effectiveness of the framework for forecasting IoV network congestion and clustering/prioritizing packets employing recurrent neural networks.

Toqa Khaled, Ahmad Al-Kabbany

We are concerned with the challenge of reliably classifying and assessing intracranial aneurysms using deep learning without compromising clinical transparency. While traditional black-box models achieve high predictive accuracy, their lack of inherent interpretability remains a significant barrier to clinical adoption and regulatory approval. Explainability is paramount in medical modeling to ensure that AI-driven diagnoses align with established neurosurgical principles. Unlike traditional eXplainable AI (XAI) methods -- such as saliency maps, which often provide post-hoc, non-causal visual correlations -- Concept Bottleneck Models (CBMs) offer a robust alternative by constraining the model's internal logic to human-understandable clinical indices. In this article, we propose an end-to-end 3D Concept Bottleneck framework that maps high-dimensional neuroimaging features to a discrete set of morphological and hemodynamic concepts for aneurysm identification. We implemented this pipeline using a pre-trained 3D ResNet-34 backbone and a 3D DenseNet-121 to extract features from CTA volumes, which were subsequently processed through a soft bottleneck layer representing human-interpretable clinical concepts. The model was optimized using a joint-loss function to balance diagnostic focal loss and concept mean squared error (MSE), validated via stratified five-fold cross-validation. Our results demonstrate a peak task classification accuracy of 93.33% +/- 4.5% for the ResNet-34 architecture and 91.43% +/- 5.8% for the DenseNet-121 model. Furthermore, the implementation of 8-pass Test-Time Augmentation (TTA) yielded a robust mean accuracy of 88.31%, ensuring diagnostic stability during inference. By maintaining an accuracy-generalization gap of less than 0.04, this framework proves that high predictive performance can be achieved without sacrificing interpretability.

Shimaa Elbana, Ahmad Kamal, Shahd Ahmed Ali et al.

The increasing size and complexity of medical imaging datasets, particularly in 3D formats, present significant barriers to collaborative research and transferability. This study investigates whether the ZFP compression technique can mitigate these challenges without compromising the performance of automated cerebrovascular segmentation, a critical first step in intracranial aneurysm detection. We apply ZFP in both its error tolerance and fixed-rate modes to a large scale, and one of the most recent, datasets in the literature, 3D medical dataset containing ground-truth vascular segmentations. The segmentation quality on the compressed volumes is rigorously compared to the uncompressed baseline (Dice approximately equals 0.8774). Our findings reveal that ZFP can achieve substantial data reduction--up to a 22.89:1 ratio in error tolerance mode--while maintaining a high degree of fidelity, with the mean Dice coefficient remaining high at 0.87656. These results demonstrate that ZFP is a viable and powerful tool for enabling more efficient and accessible research on large-scale medical datasets, fostering broader collaboration across the community.

Rojaina Mahmoud, Mona Mamdouh, Omneya Attallah et al.

In this research, we are concerned with the applicability of virtual reality-based attention training as a tool for stress management. Mental stress is a worldwide challenge that is still far from being fully managed. This has maintained a remarkable research attention on developing and validating tools for detecting and managing stress. Technology-based tools have been at the heart of these endeavors, including virtual reality (VR) technology. Nevertheless, the potential of VR lies, to a large part, in the nature of the content being consumed through such technology. In this study, we investigate the impact of a special type of content, namely, attention training, on the feasibility of using VR for stress management. On a group of fourteen undergraduate engineering students, we conducted a study in which the participants got exposed twice to a stress inducer while their EEG signals were being recorded. The first iteration involved VR-based attention training before starting the stress task while the second time did not. Using multiple features and various machine learning models, we show that VR-based attention training has consistently resulted in reducing the number of recognized stress instances in the recorded EEG signals. This research gives preliminary insights on adopting VR-based attention training for managing stress, and future studies are required to replicate the results in larger samples.

Marian K. Y. Boktor, Ahmad Al-Kabbany, Radwa Khalil et al.

This paper addresses the problem of real-time action recognition in trimmed videos, for which deep neural networks have defined the state-of-the-art performance in the recent literature. For attaining higher recognition accuracies with efficient computations, researchers have addressed the various aspects of limitations in the recognition pipeline. This includes network architecture, the number of input streams (where additional streams augment the color information), the cost function to be optimized, in addition to others. The literature has always aimed, though, at assigning the adopted network (or networks, in case of multiple streams) the task of recognizing the whole number of action classes in the dataset at hand. We propose to train multiple specialized module networks instead. Each module is trained to recognize a subset of the action classes. Towards this goal, we present a dissimilarity-based optimized procedure for distributing the action classes over the modules, which can be trained simultaneously offline. On two standard datasets--UCF-101 and HMDB-51--the proposed method demonstrates a comparable performance, that is superior in some aspects, to the state-of-the-art, and that satisfies the real-time constraint. We achieved 72.5\% accuracy on the challenging HMDB-51 dataset. By assigning fewer and unalike classes to each module network, this research paves the way to benefit from light-weight architectures without compromising recognition accuracy.