Medical Imaging Technology تكنولوجيا التصوير الطبي

Permanent URI for this collection

https://dspace.alquds.edu/handle/20.500.12213/7165

Browse

Now showing 1 - 5 of 43

Evaluation of Exposure to Electromagnetic Fields from MRI Facilities
(Al-Quds Univeersity, 2025-12-13) Noor Malek Ali Ghazawneh; نور مالك علي غزاونة
Clinical practice worldwide widely uses Magnetic Resonance Imaging (MRI) as a crucial diagnostic tool. MRI generates high-resolution images of tissues and organs by combining strong static magnetic fields (SMFs) with radiofrequency electromagnetic fields (RF-EMFs). This technique provides essential information for diagnosis, treatment planning, and patient monitoring. Although MRI is indispensable in clinical imaging, short-term exposure to strong SMFs may cause dizziness, nausea, vertigo and blurred vision. The World Health Organization (WHO) classifies long-term exposure to RF fields as possibly carcinogenic (Group 2B), while current evidence does not indicate that MRI-related SMF exposure causes cancer. These considerations indicate that there must be careful assessment and monitoring of electromagnetic field exposure in MRI facilities. This study aimed to evaluate occupational exposure to SMFs and RF electromagnetic fields in selected MRI facilities across the West Bank, Palestine. Measurements were conducted in four hospitals, which were coded for clarity and consistency throughout the study: H1 refers to the Bethlehem Arab Society for Rehabilitation (Bethlehem) operating a 1.5 T MRI scanner; H2 represents the Moscow Medical Center (Bethlehem) with a 0.55 T MRI system; H3 corresponds to the Palestine Medical Complex (Ramallah) equipped with a 1.5 T MRI scanner; and H4 denotes Al-Rahma Medical Center (Nablus), operating a 3T MRI system. RF field measurements were performed using the Spectrum Rider FPH (Rohde & Schwarz), while static magnetic fields were measured using the Spectran NF-5035 (Aaronia). Data collection followed the American College of Radiology (ACR) zoning classification and was carried out within the MRI scanner rooms and the surrounding safety zones (Zones I–IV). The results indicated that SMF levels ranged from 335 to 1100 µT across all hospitals, with the highest values recorded near the MRI scanner (Zones III–IV, ~1 m from the bore) and the lowest values observed in peripheral areas outside the facility (Zone I, ~20 m). Intermediate exposure levels were measured in patient waiting areas (Zone II, 5–8 m). All measured SMF values were well below the ICNIRP occupational exposure limit of 2 T. RF field strengths varied between 30 and 95 dBµV/m across different sequences and examinations, with higher exposure consistently observed near the bore (Zone IV) and lower levels in seating and waiting areas (Zones I–II). Central body examinations, such as pelvic, pituitary, and abdominal MRI, exhibited the highest RF exposure—particularly in Hospital H3—while peripheral examinations (knee, shoulder, and spine) showed lower levels. T2-weighted and FS/T2 sequences produced the highest RF exposure, whereas T1 and PD sequences resulted in the lowest levels. All RF exposure values remained far below the ICNIRP occupational limit of 61.4 V/m, confirming compliance with international safety standards. Regular monitoring, effective shielding, and adherence to safety protocols are recommended to maintain these safe conditions, particularly as MRI technology advances toward higher field strengths.
Efficacy of automated breast ultrasound as a screening tool for detecting breast lesions in comparison with mammography and breast biopsy
(Al-Quds Univeersity, 2024-12-14) Ahlam Mohammad Asad Mubarak; أحلام محمد مبارك
Automated breast ultrasound (ABUS) has been developed as an advanced imaging technology designed to overcome the limitations of conventional breast screening modalities, particularly the reduced sensitivity of mammography in dense breast tissue and the operator dependency of handheld ultrasound (HHUS). This study aimed to evaluate the efficacy of ABUS as a screening tool for detecting breast lesions in comparison with mammography and breast biopsy. A descriptive, retrospective cross-sectional study was conducted involving 133 women who underwent both mammography and ABUS at Yazan Radiology Center in Bethlehem between January and December 2023. The mean age of participants was 51.71 years. Breast density distribution showed that 38.3% were classified as category C, 37.6% as category B, 13.5% as category D, and 10.5% as category A. Findings demonstrated that ABUS exhibited higher sensitivity than mammography, particularly in women with dense breasts, and detected a greater number of lesions. The overall accuracy of ABUS was 61.65%, representing a statistically significant difference compared to mammography. No significant associations were observed between ABUS-detected lesions and demographic characteristics except for breast density. In conclusion, ABUS shows substantial potential as an effective screening modality, offering improved lesion detection in dense breast tissue and demonstrating advantages over mammography in sensitivity and diagnostic performance.
Deep Clustering Approaches for Carotid Artery Calcification Detection in Panoramic Radiographs for Enhancing Cardiovascular Risk Prediction
(Al-Quds University, 2026-01-08) Nadeen Khaled Ibrahim Erekat; نادين خالد إبراهيم عريقات
Cardiovascular disease remains a leading cause of death worldwide, making early identification of vascular risk markers essential for prevention. Carotid artery calcifications can occasionally be visualized on panoramic dental radiographs, offering an opportunistic indicator of atherosclerotic burden during routine dental care. Yet, manual identification is challenged by inter-reader variability and the presence of anatomical mimics, and many published AI solutions rely on supervised learning that requires large, densely labeled datasets. This thesis investigates an alternative pathway by developing an unsupervised deep clustering framework for carotid region analysis on panoramic radiographs, and by integrating questionnaire-based risk factor modeling to support broader cardiovascular risk stratification. This single-center retrospective cross-sectional observational study included 1,107 panoramic radiographs acquired between February 2025 and August 2025 during routine dental examinations at Abraj Dental Clinics affiliated with Al-Quds University, in collaboration with the Faculty of Dentistry. In addition, a cross-sectional questionnaire sub-study was prospectively administered to a subset of participants (n = 438) to capture cardiovascular risk profiles and support complementary non-imaging analyses. The imaging cohort comprised 48% males and 52% females, with an age range of 18–85 years and a mean age of approximately 40 years. Preprocessing consisted of contrast enhancement using Contrast Limited Adaptive Histogram Equalization (CLAHE) followed by extraction of bilateral carotid regions of interest (ROIs), resized to 128×128 pixels. To represent each ROI, a dual feature strategy was adopted. Interpretable handcrafted radiographic features were computed to quantify intensity distributions, texture patterns, and edge- and morphology-related cues potentially associated with calcification. In parallel, deep representations were learned using a convolutional autoencoder pretrained for 300 epochs with mean squared error loss and a low-dimensional latent space, producing compact feature vectors suitable for downstream clustering. Four clustering methods were evaluated: K-Means, hierarchical clustering, Gaussian Mixture Models, and Deep Embedded Clustering (DEC). Clustering quality was assessed using internal validation metrics that do not require ground truth labels, including the Silhouette score, Davies–Bouldin index, and Calinski–Harabasz index. To connect unsupervised clustering outcomes to clinical relevance, the thesis adopted a patient-level validation strategy, where each patient contributed left and right ROI assignments. A high-risk validation subset of 21 patients was defined using confirmed cardiovascular disease history or documented carotid calcifications in clinical records. Patient-level accuracy was calculated based on the proportion of high-risk patients grouped into the dominant high-risk cluster. Model interpretability was further supported using GradCAM++ visualization to highlight salient ROI regions consistent with expected calcification-related patterns. DEC achieved the strongest clustering performance across the internal metrics, with a Silhouette score of 0.214, a Davies–Bouldin index of 1.752, and a Calinski–Harabasz index of 524, indicating improved within-cluster cohesion and between-cluster separation relative to baseline methods. Patient-level validation also favored DEC, which achieved 95.2% accuracy in aggregating high-risk patients into a dominant cluster with fewer anomalous cases compared with K-Means, hierarchical clustering, and GMM. To incorporate non-imaging determinants of cardiovascular risk, prospectively administered questionnaire data were analyzed for 438 participants (220 high risk, 218 low risk). Univariate association testing using chi-square statistics with Cramér’s V suggested notable associations with age, physical activity, sleep duration, sedentary time, dietary behaviors, selected health awareness indicators, and mental health measures such as relaxation difficulty and feelings of worthlessness. A Random Forest classifier trained on questionnaire features achieved high predictive performance (accuracy = 0.9318, ROC AUC = 0.9821, F1 = 0.9291), and feature importance analysis highlighted socioeconomic factors and psychological distress-related variables among influential predictors alongside age and lifestyle behaviors. In conclusion, this thesis demonstrates that unsupervised deep clustering of anatomically defined carotid ROIs on panoramic radiographs can yield coherent groupings aligned with clinically defined high-risk status, while offering an interpretable and label-efficient screening support pathway. Combining imaging-based signals with questionnaire-derived risk factors may further strengthen cardiovascular risk stratification and support targeted referral for confirmatory vascular assessment. Future work should incorporate gold-standard vascular confirmation, multi-site external validation, and longitudinal outcome linkage to establish clinical reliability and generalizability.
Cross-Modality Transfer Learning for Reliable Lung Cancer Nodule Classification in Low-Dose CT
(Al-Quds University, 2026-01-08) Sara Rasheed Saleh Asfour; سارة عصفور
Lung cancer is the leading cause of cancer-related mortality worldwide and remains a major health burden in Palestine. According to the Palestine Annual Health Report 2024, lung and bronchus cancers accounted for 316 newly diagnosed cases (10.5 per 100,000 population) and 266 deaths, with 85% occurring among males. These epidemiological patterns highlight the urgent need for locally validated AI solutions that support early lung cancer detection and robust clinical decision-making, particularly in resource-limited healthcare settings where radiological expertise and structured screening programs remain limited. This study proposes a dose-aware hybrid deep-learning–machine-learning framework for multi-class Lung-RADS classification using heterogeneous chest CT datasets collected from multiple Palestinian institutions. The framework integrates optimized contrast enhancement, deep feature extraction using VGG16, and five classical machine-learning classifiers (LR, SVM, RF, GB, and DT). Contrast enhancement was systematically evaluated using two subsets of 120 malignant cases, one LDCT and one SDCT which is demonstrating that the hybrid CLAHE-USM provided the most balanced improvements (EME = 20.648, PSNR = 19.711, SSIM = 0.912). Dose-optimized parameters, including a clip limit of 3 for LDCT and 4 for SDCT, confirmed the importance of dose-specific preprocessing in stabilizing image quality. The classification results obtained across the LDCT, combined, external validation, and clinical validation datasets demonstrate that the proposed VGG16-machine learning framework provides robust and consistent performance for Lung-RADS-based nodule risk stratification. On the LDCT dataset, all classifiers achieved clinically meaningful performance, although variability was observed across Lung-RADS categories. GB and decision tree classifiers demonstrated powerful performance in LR2 and higher-risk categories, achieving accuracies above 0.85. A clear performance improvement was observed with increasing risk of malignancy. For LR4A and LR4B, most classifiers achieved higher accuracy than in lower-risk categories, reflecting the greater structural distinctiveness of high-risk nodules. The most important performance gains were observed on the combined LDCT-SDCT dataset, particularly for LR3 and LR4A-LR4B. In this setting, SVM and RF achieved near-ceiling performance, with accuracies exceeding 0.97 and AUC values approaching unity. External validation on an independent test set from Al-Makassed Hospital further confirmed the robustness of the proposed framework. For LR2, all major classifiers achieved accuracies above 0.84, whereas for LR4B, all major classifiers achieved accuracies above 0.86. The most stringent assessment of diagnostic performance was provided by Clinical validation on biopsy-confirmed LR4B cases from Augusta Victoria Hospital. SVM achieved the highest accuracy and AUC, followed closely by LR. Overall, this framework demonstrates strong potential as a practical decision-support tool for improving Lung-RADS-based risk stratification and supporting early lung cancer detection in resource-constrained healthcare settings
A Systematic Comparative Evaluation of Classical Machine Learning Algorithms for Liver Lesion Classification in Ultrasound Imaging
(Al- Quds University, 2026-01-14) Aesha Loay Ebrahim Enairat; عائشه لؤي ابراهيم انعيرات
Liver ultrasound is very popular as it is accessible, safe and inexpensive, but the distinction between benign and malignant liver lesions is still difficult to make, owing to the noise, low contrast, and the overlap of the lesions. The thesis is an evaluation of the performance and generalization of classical machine learning techniques in liver lesion classification with ultrasound images in 3 binary tasks, where one of them is benign-normal, another one is malignant-normal, and the last one is benign-malignant. We merged a localized clinical dataset with a publicly available dataset in Zenodo, which created an original set of 6,791 ultrasound images. By eliminating the duplicate and very similar images to avoid redundancy we were left with a curated set of unique images numbered 2,387. We compared the default preprocessing with contrast enhancement with Contrast Limited Adaptive Histogram Equalization (CLAHE) and tested various traditional classifiers, including ensemble model, support vector machines, linear model, instance-based model as well as probabilistic models. Stratified cross-validation, independent testing and a fully isolated holdout set were used in the evaluation of model performance. The best and consistent performance of the ensemble-based classifiers was observed especially when it came to malignant- normal and benign- normal classification. Conversely, benign-malignant classification was the hardest to carry out because the lesion types had large visual overlaps. The use of CLAHE resulted in better sensitivity and lesion separability of a number of models, and task-specific advantages. On the whole, the findings suggest that classical machine learning pipelines with the proper support of preprocessing and strict validation may be effective in terms of assisting the classification of ultrasound-based liver lesions and may serve as the basis of further advancement.

Browse

Recent Submissions