Razan AbuGhoush

Mrs. Razan AbuGhoush

orcid

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
29 May 2025 in Children
Verified
Journal Article
Open access
PDF file
STSuliman Thwib
Suliman Thwib
RQRadwan Qasrawi
Radwan Qasrawi
GIGhada Issa
Ghada Issa
MAMalak Amro
Malak Amro
RARazan AbuGhoush
Razan AbuGhoush
MNMaisan Nimer
Maisan Nimer
Background: The growth patterns of children and adolescents are influenced by multiple factors. This study employed structural equation modeling (SEM) to determine the primary factors influencing the growth of Palestinian children and adolescents in the West Bank (WB). Methods: A cross-sectional survey conducted in 2022 in the WB collected data from 1400 households, of which 500 with children aged 5–17 years and were selected for analysis. The survey assessed household food insecurity, socioeconomic status (SES), parental health history, nutritional awareness, food consumption patterns, and nutrient intake. The latent construct “Child Growth” was measured using Body Mass Index-for-age Z-score (BAZ), height-for-age Z-score (HAZ), and Mean Adequacy Ratio (MAR). SEM was employed to examine the interrelations among latent variables and their observed indicators. Results: Mineral intake showed the strongest direct effect on child growth (β = 0.812, p < 0.001), followed by food availability (β = 0.492), vitamin intake (β = 0.481), SES (β = 0.439), and macronutrient intake (β = 0.383). MAR exhibited the highest factor loading as a growth indicator, while HAZ had limited significance, suggesting its chronic nature. SES influenced growth both directly and indirectly through its effects on food availability and dietary intake pathways. Strong correlations between nutritional knowledge and nutrient classes reflect the interrelationship between behavioral and biological determinants. Conclusions: Both structural and immediate nutritional factors influence child growth. In Palestinian youth, mineral intake, food security, and SES have the greatest influence. These findings offer a framework for understanding the growth determinants of Palestinian youth in the WB and support the development of targeted interventions to improve dietary quality and overall nutritional status.

8 September 2025 in PLOS ONE
Verified
Journal Article
Open access
PDF file
GIGhada Issa
Ghada Issa
RQRadwan Qasrawi
Radwan Qasrawi
STSuliman Thwib
Suliman Thwib
MAMalak Amro
Malak Amro
RARazan AbuGhoush
Razan AbuGhoush
MNMaisan Nimer
Maisan Nimer
Background: Undernutrition remains a persistent public health concern among young children in Palestine, shaped by a range of socioeconomic and dietary factors. This study applies a Structural Equation Modeling (SEM) approach to explore both direct and indirect determinants of child growth among children aged 6–59 months in the West Bank. Methods: Data were drawn from a 2022 cross-sectional survey involving 300 children selected from 1,400 households. Child growth was assessed using anthropometric indicators (HAZ, WAZ, MUAC) and dietary adequacy via the Mean Nutrient Inadequacy Ratio (MNIR). Latent constructs were developed for socioeconomic status, food insecurity, dietary intake (macro- and micronutrients), feeding practices, neonatal health, and food availability. SEM was used to examine interrelationships and the pathways influencing child nutritional status. Results: Descriptive analysis showed substantial variation in undernutrition and nutrient intake across regions, family size, and parental education. SEM results showed that macronutrient intake had the strongest direct effect on child growth, followed by socioeconomic status and neonatal health. Infant feeding practices and micronutrient intake also contributed positively, though their effects were primarily indirect. Food availability acted as a significant mediator, linking structural access issues to poor dietary diversity and feeding behavior. The model demonstrated good fit across multiple indices. Conclusion: Child nutrition in Palestine is shaped by interrelated determinants of dietary intake, caregiver practices, food access, and early-life health conditions. Effective interventions must target both nutritional quality and the broader social and environmental context to improve growth outcomes. SEM offers a powerful framework to understand these pathways and guide evidence-based policy design.

28 November 2025 in Journal of Imaging
Verified
Journal Article
Open access
PDF file
RQRadwan Qasrawi
Radwan Qasrawi
STSuliman Thwib
Suliman Thwib
GIGhada Issa
Ghada Issa
Ibraheem Bassam QdaihIbraheem Bassam Qdaih
Ibraheem Bassam Qdaih
RARazan AbuGhoush
Razan AbuGhoush
HAHamza Arjah
Hamza Arjah
Background: The early and accurate classification of ischemic stroke stages on computed tomography (CT) remains challenging due to subtle attenuation differences and significant scanner variability. This study developed a neural network framework to dynamically optimize Hounsfield Unit (HU) transformations and CLAHE parameters for temporal stage-specific stroke classification. Methods: We analyzed 1480 CT cases from 68 patients across five stages (hyperacute, acute, subacute, chronic, and normal). The training data were augmented via horizontal flipping, ±7° rotation. A convolutional neural network (CNN) was used to optimize linear transformation and CLAHE parameters through a combined loss function incorporating the effective measure of enhancement (EME), peak signal-to-noise ratio (PSNR), and regularization. the enhanced images were classified using logistic regression (LR), support vector machines (SVMs), and random forests (RFs) with 25-fold cross-validation. Model interpretability was evaluated using Grad-CAM. Results: Neural network optimization significantly outperformed static parameters across validation metrics. Deep CLAHE achieved the following accuracies versus static CLAHE: hyperacute (0.9838 vs. 0.9754), acute (0.9904 vs. 0.9873), subacute (0.9948 vs. 0.9825), and chronic (near-perfect 0.9979 vs. 0.9808). Qualitative interpretability analysis confirmed that models focused on clinically relevant regions, with optimized enhancement producing more coherent attention patterns than static methods. Parameter analysis revealed stage-aware adaptation: conservative enhancement in early phases (slope: 1.249–1.257), maximized in subacute (slope: 1.290–1.292), and restrained in the chronic phase (slope: 1.240–1.258), reflecting underlying stroke pathophysiology. Conclusions: A neural network-optimized framework with interpretability validation provides stage-specific stroke classification that achieves superior performance over static methods. Its pathophysiology-aligned parameter adaptation offers a clinically viable and transparent solution for emergency stroke assessment.

23 July 2026 in PLOS ONE
Verified
Journal Article
Open access
PDF file
RQRadwan Qasrawi
Radwan Qasrawi
STSuliman Thwaib
Suliman Thwaib
GIGhada Issa
Ghada Issa
RTRand Al Taweel
Rand Al Taweel
RARazan AbuGhoush
Razan AbuGhoush
AJAyoub Jawaldeh
Ayoub Jawaldeh
Background: Adults with disabilities experience increased cardiometabolic risk, yet the pathways linking nutrition, functional ability, and environmental factors remain insufficiently defined. This study examined these relationships using a structural equation modeling framework. Methods: A cross-sectional study was conducted among 123 adults with disabilities in Palestine (mean age 31.9 years). Data included anthropometric measures, two non‑consecutive 24-hour dietary recalls, disability severity (Disability Rating Scale), functional ability (Barthel Index), psychosocial environment, and socioeconomic indicators. Dietary quality and adequacy were assessed using the Diet Quality Index (DQI) and Mean Adequacy Ratio (MAR). Seven latent constructs were specified and tested using confirmatory factor analysis and structural equation modeling. Results: Indicators of cardiometabolic risk were prevalent: 58.5% overweight/obese and 51.2% had inadequate intake. Disability severity correlated with reduced muscle mass and poorer diet quality. Nutritional adequacy positively predicted body composition (β = 0.274) and functional ability (β = 0.185). Functional ability and physical activity were inversely associated with adiposity, while sedentary behavior moderated the relationship between diet quality and waist circumference. Mental illness specifically impaired diet quality, but not caloric intake. Urban residence and lower education predicted poorer dietary outcomes. The model explained 63.7% of variance in body composition and 94.4% in functional ability. Conclusions: Cardiometabolic risk in adults with disabilities is shaped by interrelated pathways involving disability severity, nutrition, functional capacity, and psychosocial factors. Lower anthropometric measures may reflect muscle loss rather than reduced risk. Interventions should focus on improving nutritional adequacy, supporting functional independence, addressing mental health, and reducing environmental barriers.

21 September 2026 in Journal of Imaging
Verified
Journal Article
Open access
PDF file
RQRadwan Qasrawi
Radwan Qasrawi
RARazan AbuGhoush
Razan AbuGhoush
GIGhada Issa
Ghada Issa
STSuliman Thwib
Suliman Thwib
MAMalak Amro
Malak Amro
MQMarwan Qubja
Marwan Qubja
Automated CT lung nodule classification suffers from domain shift across centers. We propose LoRA-GRL, combining Low-Rank Adaptation (LoRA) with adversarial domain harmonization via a Gradient Reversal Layer. Rank-8 LoRA modules were inserted into all attention and feed-forward linear projections (48 sites) of a frozen ViT-S backbone; a domain discriminator encouraged center-invariant features across three centers. Normal vs. Benign and Normal vs. Malignant tasks used 264 patients with patient-level five-fold stratified cross-validation. LoRA-GRL achieved AUCs of 0.9735 and 0.9869, close to full fine-tuning (0.9747 and 0.9863). Differences from strongest baselines fell within overlapping 95% CIs under paired bootstrap, indicating comparable discrimination, not superiority. Efficiency is the main advantage: only 0.789 M trainable parameters, a 96.4% reduction from 21.865 M, and inference latency matched full fine-tuning after adapter merging. Grad-CAM showed nodule-localized predictions but also slice-selection failures. For benign classification, higher AUC than plain LoRA came with higher specificity but lower sensitivity. For malignant classification, sensitivity was 0.9412 (5.9 pp above full fine-tuning) and specificity 0.9697. The small cross-center AUC range (0.0011) reflects internal consistency across participating centers, not unseen-site generalization. LoRA-GRL is a promising parameter-efficient candidate for multi-center lung nodule classification, potentially reducing scanner-specific bias and storage/memory needs, but clinical utility requires external validation and prospective evaluation.

Verified
Journal Article
Open access
PDF file
STSuliman Thwib
Suliman Thwib
RQRadwan Qasrawi
Radwan Qasrawi
GIGhada Issa
Ghada Issa
RARazan AbuGhoush
Razan AbuGhoush
HAHUSSEIN ALMASRI
HUSSEIN ALMASRI
MQMarah Qawasmi
Marah Qawasmi
Background: Real-time and accurate analysis of breast ultrasounds is crucial for diagnosis but remains challenging due to issues like low image contrast and operator dependency. This study aims to address these challenges by developing an integrated framework for real-time lesion detection and tracking. Methods: The proposed system combines Contrast-Limited Adaptive Histogram Equalization (CLAHE) for image preprocessing, a transfer learning-enhanced YOLOv11 model following a continual learning paradigm for cross-center generalization in for lesion detection, and a novel Detection-Based Tracking (DBT) approach that integrates Kernelized Correlation Filters (KCF) with periodic detection verification. The framework was evaluated on a dataset comprising 11,383 static images and 40 ultrasound video sequences, with a subset verified through biopsy and the remainder annotated by two radiologists based on radiological reports. Results: The proposed framework demonstrated high performance across all components. The transfer learning strategy (TL12) significantly improved detection outcomes, achieving a mean Average Precision (mAP) of 0.955, a sensitivity of 0.938, and an F1 score of 0.956. The DBT method (KCF + YOLO) achieved high tracking accuracy, with a success rate of 0.984, an Intersection over Union (IoU) of 0.85, and real-time operation at 54 frames per second (FPS) with a latency of 7.74 ms. The use of CLAHE preprocessing was shown to be a critical factor in improving both detection and tracking stability across diverse imaging conditions. Conclusions: This research presents a robust, fully integrated framework that bridges the gap between speed and accuracy in breast ultrasound analysis. The system’s high performance and real-time efficiency underscore its strong potential for clinical adoption to enhance diagnostic workflows, reduce operator variability, and improve breast cancer assessment.