Saied Salem

Saied Salem

Auto-generated profile, based on publicly available information

Saied Salem

Department of Artificial Intelligence, College of AI Convergence, Daeyang AI Center, Sejong University, Seoul, Republic of Korea

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

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

Representative Publications

View All Publications
3 May 2025 in Artificial Intelligence Review
Verified
Journal Article
Open access
PDF file
Mugahed A. Al-antariMugahed A. Al-antari
Mugahed A. Al-antari
SSSaied Salem
Saied Salem
MRMukhlis Raza
Mukhlis Raza
AEAhmed S. Elbadawy
Ahmed S. Elbadawy
EBErtan Bütün
Ertan Bütün
YGYeonghyeon Gu
Yeonghyeon Gu
Lumbar spinal stenosis (LSS) involves the narrowing of the spinal canal, leading to compression of the spinal cord and nerves in the lower back. Common causes include injuries, degenerative age-related changes, congenital conditions, and tumors, all of which contribute to back pain. Early diagnosis is critical for symptom management, preventing progression, and preserving quality of life. This study systematically reviews AI-based approaches for predicting LSS using MRI axial and sagittal imaging. The review focuses on various AI tasks: detection, segmentation, classification, hybrid approaches, spinal index measurements (SIM), and explainable AI frameworks. The aim is to highlight current knowledge, identify limitations in existing models, and propose future research directions. Following PRISMA guidelines and the PICO method (Population, Intervention, Comparison, Outcome), the review collects data from databases like PubMed, Web of Science, ScienceDirect, and IEEE Xplore (2005–2024). The Rayyan AI tool is used for duplicate removal and screening. The screening process includes an initial review of titles and abstracts, followed by full-text appraisal. The Meta Quality Appraisal Tool (MetaQAT) assesses the quality of selected articles. Of 1323 records, 97 duplicates were removed. After screening, 895 records were excluded, leaving 331 for full-text review. Among these, 184 articles were excluded for lacking AI relevance. Ultimately, 95 key articles (91 technical papers and 4 reviews) were identified for their contributions to AI-based LSS prediction. This review provides a comprehensive analysis of AI techniques in LSS prediction, guiding future research and advancing understanding in areas like explainable AI and large language models (LLMs).

6 September 2025
Verified
Conference Paper
Closed access
HKHyunwook Kwon
Hyunwook Kwon
SSSaied Salem
Saied Salem
MRMukhlis Raza
Mukhlis Raza
AHAfnan Habib
Afnan Habib
EBErtan Bütün
Ertan Bütün
Mugahed A. Al-antariMugahed A. Al-antari
Mugahed A. Al-antari
Lumbar spinal disorders such as stenosis and degeneration are common and debilitating. While lumbar spine MRI is critical for diagnosis, conventional assessments rely on subjective visual grading. To overcome this, we propose a deep learning-based segmentation framework that automatically extracts anatomical landmarks and measures clinically relevant indicators from both sagittal and axial MRI. Our geometry-guided pipeline quantifies vertebral and disc heights, foraminal distances, and spinal canal dimensions with explainable outputs. The system achieved high agreement with expert measurements (MAE $<1.3 ~\text{mm}, \mathrm{r}=0.98$), and outlier filtering improved overall coverage to over 94 %. Designed for transparency and adaptability, the framework offers potential for application to other spinal conditions such as scoliosis and postoperative instability. This work moves toward trustworthy, quantitative AI tools for clinical spinal analysis.

28 January 2026 in IEEE Access
Verified
Journal Article
Open access
QYQuan Yuan
Quan Yuan
HKHyunwook Kwon
Hyunwook Kwon
MRMukhlis Raza
Mukhlis Raza
SLSoojeong Lee
Soojeong Lee
AAAli Alqahtani
Ali Alqahtani
Mugahed A. Al-antariMugahed A. Al-antari
Mugahed A. Al-antari
Vehicle detection plays a vital role in a variety of applications such as autonomous driving, traffic management, and public safety. In this article, we propose an ensemble artificial intelligence framework to simultaneously detect and classify various vehicles on public roads. The framework aims to improve the detection accuracy of vehicles and their small targets at long distances in different environmental conditions: rainy, foggy, and night scenes. To achieve this goal, we first construct a benchmark vehicle dataset (Dataset 4) containing multiple vehicle types (Truck, Van, Car, Bus and Two-wheelers) across complex environmental conditions. Subsequently, we adopt multiple state-of-the-art AI detectors, including finetuned versions of YOLOv5, YOLOv7, YOLOv8, YOLOv10, DETR-ResNet50, and Faster R-CNN. Using the random-search approach the hyperparameters of the AI models are automatically optimized to select and fuse the top three models into a unified detection framework through an ensemble learning strategy. The ensemble framework aims to combine the advantages of multiple detectors to improve the detection accuracy and robustness under different vehicle types and conditions. Our proposed integrated model framework based on YOLOv7, YOLOv8 and YOLOv10 further improves the model performance using non-maximum suppression (NMS) and dynamic weighting methods, outperforming single detectors with excellent detection performance of 94.2% precision, 94.0% recall, 94.1% F1-score, and 96.6% mAP. The main contribution is the proposal of a novel vehicle detection integration framework and a comprehensive benchmark dataset to address the challenges faced by current vehicle detection systems. The system is deployed to test real driving videos, and the experimental results demonstrate the practical applicability of our proposed method.

6 September 2025
Verified
Conference Paper
Closed access
SSSaied Salem
Saied Salem
AHAfnan Habib
Afnan Habib
MRMukhlis Raza
Mukhlis Raza
AAAhmet Arif Aydin
Ahmet Arif Aydin
YGYeonghyeon Gu
Yeonghyeon Gu
Mugahed A. Al-antariMugahed A. Al-antari
Mugahed A. Al-antari
Automated analysis of lumbar spine MRI is essential for improving diagnostic consistency and enhancing clinical workflow efficiency in the evaluation of chronic low back pain (CLBP). In this study, we present a computer-aided diagnosis (CAD) framework designed to automate both the analysis of lumbar spine MRI scans and the generation of structured diagnostic reports. The system processes 3D DICOM MRI volumes by extracting mid-sagittal slices for the segmentation of vertebrae and intervertebral discs (IVDs), followed by a 3D cross-projection method to localize the corresponding axial slices. The SegResNet architecture is employed as segmentation model to delineate anatomical structures in both sagittal and axial views. From these segmentations, quantitative measurements of key spinal anatomy are extracted, enabling automated anatomical indices measurements, disorders detection, spinal stenosis severity grading, and evaluation of spinal alignment abnormalities. These assessments serve as the diagnostics information feed into the input prompt to large language model (LLM) for report generation. The proposed system leverages a novel retrievalaugmented generation (RAG) approach that integrates semantic retrieval and knowledge graph-based reasoning to generate detailed, level-specific diagnostic report. The system demonstrates high segmentation performance (Dice: 97.79% sagittal, 93.52% axial) and generates clinically coherent reports using AgenticRAG, achieving a BERT F1-score of 83.58%. These results highlight its effectiveness for accurate, level-specific diagnosis and streamlined clinical reporting.