{"id":{"repo_id":"csusb","oai_identifier":"oai:scholarworks.lib.csusb.edu:etd-3226"},"canonical_url":"https://search.ndltd.org/etd/csusb/oai:scholarworks.lib.csusb.edu:etd-3226","repository":{"repo_id":"csusb","name":"CSUniversity San Bernardino","base_url":"https://scholarworks.lib.csusb.edu/do/oai/"},"display":{"title":"OSTEOPOROSIS DISEASE DETECTION USING HYBRID DEEP LEARNING","abstract":"Osteoporosis is a silently progressing skeletal disorder characterized by deteriorating bone mass and density, leading to increased fragility and a heightened risk of fractures. Early identification and treatment are pivotal in preventing osteoporotic fractures, which are associated with significant morbidity and mortality. However, the conventional methods for osteoporosis detection, such as dual-energy X-ray absorptiometry, are constrained by accessibility, cost, and radiation exposure. This thesis introduces a novel hybrid deep learning model that synergistically integrates the architectural strengths of VGG19 and DenseNet-121 for the enhanced detection and classification of osteoporosis from X-ray images. The comprehensive utilization of the deep feature extraction capabilities of VGG19, with its sequential convolutional framework, and the parameter-efficient, densely connected pathways of DenseNet-121, provides an advanced approach to medical image analysis. The methodology involves the use of a curated dataset of X-ray images that have been preprocessed to optimize the image quality for feature learning. The pretrained models of VGG19 and DenseNet-121 initially trained on extensive datasets are fine-tuned on this dataset, with layers from both networks merged to form the hybrid architecture. This fine-tuning allows for adaptation to the specific textural and morphological patterns indicative of osteoporosis. Innovations in the image preprocessing with data augmentation, combined with meticulous hyperparameter optimization, underpin the accurate classification of bone densities. The performance of various deep learning models for osteoporosis detection was evaluated and compared. Single architecture models, including VGG19 and DenseNet-121, achieved accuracies ranging from 63% to 66% on the test dataset, regardless of the optimizer used or the number of training epochs. Notably, the proposed hybrid deep learning model, which synergistically combines VGG19 and DenseNet-121 architectures, significantly outperformed individual models, achieving a remarkable accuracy of 95.83% after only 10 epochs using the Adam optimizer. This substantial improvement in accuracy demonstrates the efficacy of the hybrid approach in capturing complex features indicative of osteoporosis from X-ray images, potentially offering a more reliable tool for early detection and diagnosis of the disease","abstract_html":"Osteoporosis is a silently progressing skeletal disorder characterized by deteriorating bone mass and density, leading to increased fragility and a heightened risk of fractures. Early identification and treatment are pivotal in preventing osteoporotic fractures, which are associated with significant morbidity and mortality. However, the conventional methods for osteoporosis detection, such as dual-energy X-ray absorptiometry, are constrained by accessibility, cost, and radiation exposure. This thesis introduces a novel hybrid deep learning model that synergistically integrates the architectural strengths of VGG19 and DenseNet-121 for the enhanced detection and classification of osteoporosis from X-ray images. The comprehensive utilization of the deep feature extraction capabilities of VGG19, with its sequential convolutional framework, and the parameter-efficient, densely connected pathways of DenseNet-121, provides an advanced approach to medical image analysis. The methodology involves the use of a curated dataset of X-ray images that have been preprocessed to optimize the image quality for feature learning. The pretrained models of VGG19 and DenseNet-121 initially trained on extensive datasets are fine-tuned on this dataset, with layers from both networks merged to form the hybrid architecture. This fine-tuning allows for adaptation to the specific textural and morphological patterns indicative of osteoporosis. Innovations in the image preprocessing with data augmentation, combined with meticulous hyperparameter optimization, underpin the accurate classification of bone densities. The performance of various deep learning models for osteoporosis detection was evaluated and compared. Single architecture models, including VGG19 and DenseNet-121, achieved accuracies ranging from 63% to 66% on the test dataset, regardless of the optimizer used or the number of training epochs. Notably, the proposed hybrid deep learning model, which synergistically combines VGG19 and DenseNet-121 architectures, significantly outperformed individual models, achieving a remarkable accuracy of 95.83% after only 10 epochs using the Adam optimizer. This substantial improvement in accuracy demonstrates the efficacy of the hybrid approach in capturing complex features indicative of osteoporosis from X-ray images, potentially offering a more reliable tool for early detection and diagnosis of the disease","abstract_has_math":false,"creators":["Manchanayakanahalli Basavaraju, Pragathi"],"institution":null,"degree_name":"Master of Science in Computer Science","degree_level":"Restricted Project: Campus only access","degree_discipline":"School of Computer Science and Engineering","degree_department":null,"school":null,"contributors":["Muheidat, Fadi"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":null,"date_issued":"","date_published":null,"updated_at":"2026-11-19T08:09:50Z","subjects":["OSTEOPOROSIS DISEASE DETECTION","Computer Engineering"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[],"person_identifier_entries":[]},"links":{"outbound_url":"https://scholarworks.lib.csusb.edu/etd/2047","outbound_label":"Repository record","outbound_source":"dc:identifier","record_url":"https://scholarworks.lib.csusb.edu/etd/2047","record_label":"Repository record","record_kind":"record","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://scholarworks.lib.csusb.edu/do/oai/?verb=GetRecord&metadataPrefix=oai_etdms&identifier=oai%3Ascholarworks.lib.csusb.edu%3Aetd-3226","prefix":"oai_etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Muheidat, Fadi"]},{"key":"dc:creator","label":"Author","values":["Manchanayakanahalli Basavaraju, Pragathi"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.available","label":"Dc Date Available","values":["2024-10-30T07:00:00Z"]},{"key":"dc:date.created","label":"Dc Date Created","values":["2024-12-01T08:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["School of Computer Science and Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Restricted Project: Campus only access"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Science in Computer Science"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["OSTEOPOROSIS DISEASE DETECTION","Computer Engineering"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://scholarworks.lib.csusb.edu/etd/2047"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Osteoporosis is a silently progressing skeletal disorder characterized by deteriorating bone mass and density, leading to increased fragility and a heightened risk of fractures. Early identification and treatment are pivotal in preventing osteoporotic fractures, which are associated with significant morbidity and mortality. However, the conventional methods for osteoporosis detection, such as dual-energy X-ray absorptiometry, are constrained by accessibility, cost, and radiation exposure. This thesis introduces a novel hybrid deep learning model that synergistically integrates the architectural strengths of VGG19 and DenseNet-121 for the enhanced detection and classification of osteoporosis from X-ray images. The comprehensive utilization of the deep feature extraction capabilities of VGG19, with its sequential convolutional framework, and the parameter-efficient, densely connected pathways of DenseNet-121, provides an advanced approach to medical image analysis. The methodology involves the use of a curated dataset of X-ray images that have been preprocessed to optimize the image quality for feature learning. The pretrained models of VGG19 and DenseNet-121 initially trained on extensive datasets are fine-tuned on this dataset, with layers from both networks merged to form the hybrid architecture. This fine-tuning allows for adaptation to the specific textural and morphological patterns indicative of osteoporosis. Innovations in the image preprocessing with data augmentation, combined with meticulous hyperparameter optimization, underpin the accurate classification of bone densities. The performance of various deep learning models for osteoporosis detection was evaluated and compared. Single architecture models, including VGG19 and DenseNet-121, achieved accuracies ranging from 63% to 66% on the test dataset, regardless of the optimizer used or the number of training epochs. Notably, the proposed hybrid deep learning model, which synergistically combines VGG19 and DenseNet-121 architectures, significantly outperformed individual models, achieving a remarkable accuracy of 95.83% after only 10 epochs using the Adam optimizer. This substantial improvement in accuracy demonstrates the efficacy of the hybrid approach in capturing complex features indicative of osteoporosis from X-ray images, potentially offering a more reliable tool for early detection and diagnosis of the disease"]},{"key":"dc:title","label":"Title","values":["OSTEOPOROSIS DISEASE DETECTION USING HYBRID DEEP LEARNING"]}]}],"canonical_facts":{"dc:contributor":["Muheidat, Fadi"],"dc:creator":["Manchanayakanahalli Basavaraju, Pragathi"],"dc:date.available":["2024-10-30T07:00:00Z"],"dc:date.created":["2024-12-01T08:00:00Z"],"dc:description.abstract":["Osteoporosis is a silently progressing skeletal disorder characterized by deteriorating bone mass and density, leading to increased fragility and a heightened risk of fractures. Early identification and treatment are pivotal in preventing osteoporotic fractures, which are associated with significant morbidity and mortality. However, the conventional methods for osteoporosis detection, such as dual-energy X-ray absorptiometry, are constrained by accessibility, cost, and radiation exposure. This thesis introduces a novel hybrid deep learning model that synergistically integrates the architectural strengths of VGG19 and DenseNet-121 for the enhanced detection and classification of osteoporosis from X-ray images. The comprehensive utilization of the deep feature extraction capabilities of VGG19, with its sequential convolutional framework, and the parameter-efficient, densely connected pathways of DenseNet-121, provides an advanced approach to medical image analysis. The methodology involves the use of a curated dataset of X-ray images that have been preprocessed to optimize the image quality for feature learning. The pretrained models of VGG19 and DenseNet-121 initially trained on extensive datasets are fine-tuned on this dataset, with layers from both networks merged to form the hybrid architecture. This fine-tuning allows for adaptation to the specific textural and morphological patterns indicative of osteoporosis. Innovations in the image preprocessing with data augmentation, combined with meticulous hyperparameter optimization, underpin the accurate classification of bone densities. The performance of various deep learning models for osteoporosis detection was evaluated and compared. Single architecture models, including VGG19 and DenseNet-121, achieved accuracies ranging from 63% to 66% on the test dataset, regardless of the optimizer used or the number of training epochs. Notably, the proposed hybrid deep learning model, which synergistically combines VGG19 and DenseNet-121 architectures, significantly outperformed individual models, achieving a remarkable accuracy of 95.83% after only 10 epochs using the Adam optimizer. This substantial improvement in accuracy demonstrates the efficacy of the hybrid approach in capturing complex features indicative of osteoporosis from X-ray images, potentially offering a more reliable tool for early detection and diagnosis of the disease"],"dc:identifier":["https://scholarworks.lib.csusb.edu/etd/2047"],"dc:subject":["OSTEOPOROSIS DISEASE DETECTION","Computer Engineering"],"dc:title":["OSTEOPOROSIS DISEASE DETECTION USING HYBRID DEEP LEARNING"],"thesis:degree_discipline":["School of Computer Science and Engineering"],"thesis:degree_level":["Restricted Project: Campus only access"],"thesis:degree_name":["Master of Science in Computer Science"]},"updated_at":"2026-11-19T08:09:50Z"}