Pulmonary fibrosis is a progressive interstitial lung disease characterized by the accumulation of fibrotic tissue within the lungs, leading to impaired respiratory function and reduced quality of life. Early detection is important for disease management; however, accurate diagnosis often relies on high-resolution computed tomography (CT), which may not be accessible in all clinical settings. Chest radiography provides a lower-cost and widely available imaging modality, but interpretation of chest X-rays for fibrotic disease can be challenging due to subtle radiographic patterns and overlapping anatomical structures. This thesis investigates the use of multimodal deep learning techniques to assist in pul- monary fibrosis detection by integrating chest X-ray images with corresponding radiology reports. Pretrained medical vision–language models, including BioViL-T, PubMedCLIP, and KAD, are used to generate embedding representations for both imaging and textual modalities. These embeddings are combined through a multimodal feature representation and processed by a neural network classifier to predict the presence of pulmonary fibro- sis. The proposed framework is evaluated using the PadChest dataset, which contains chest radiographs paired with radiology reports. Experimental results demonstrate that incorporating textual radiology reports alongside imaging features can improve diagnostic performance compared to models that rely solely on image-based representations. The findings highlight the potential of multimodal learning frameworks to better capture clinically relevant information by integrating visual imaging patterns with expert textual interpretation. This work contributes to the development of computational tools designed to support clinicians in pulmonary fibrosis assessment using widely available chest radiography data.
","abstract_html":"<p>Pulmonary fibrosis is a progressive interstitial lung disease characterized by the accumulation of fibrotic tissue within the lungs, leading to impaired respiratory function and reduced quality of life. Early detection is important for disease management; however, accurate diagnosis often relies on high-resolution computed tomography (CT), which may not be accessible in all clinical settings. Chest radiography provides a lower-cost and widely available imaging modality, but interpretation of chest X-rays for fibrotic disease can be challenging due to subtle radiographic patterns and overlapping anatomical structures. This thesis investigates the use of multimodal deep learning techniques to assist in pul- monary fibrosis detection by integrating chest X-ray images with corresponding radiology reports. Pretrained medical vision–language models, including BioViL-T, PubMedCLIP, and KAD, are used to generate embedding representations for both imaging and textual modalities. These embeddings are combined through a multimodal feature representation and processed by a neural network classifier to predict the presence of pulmonary fibro- sis. The proposed framework is evaluated using the PadChest dataset, which contains chest radiographs paired with radiology reports. Experimental results demonstrate that incorporating textual radiology reports alongside imaging features can improve diagnostic performance compared to models that rely solely on image-based representations. The findings highlight the potential of multimodal learning frameworks to better capture clinically relevant information by integrating visual imaging patterns with expert textual interpretation. This work contributes to the development of computational tools designed to support clinicians in pulmonary fibrosis assessment using widely available chest radiography data.</p>","abstract_has_math":false,"creators":["Noel-Rickert, Adele J"],"institution":null,"degree_name":"Master of Science","degree_level":"Thesis","degree_discipline":"Math and Computer Science","degree_department":null,"school":null,"contributors":["Vinay Shashidhar"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026-04-01T07:00:00Z","date_published":"2026-04-01T07:00:00Z","updated_at":"2026-07-24T03:24:38Z","subjects":["Pulmonary fibrosis","Chest X-ray","Multimodal deep learning","Medical imaging","Vision-language models","Radiology reports","Deep learning","Medical image analysis","Computer-aided diagnosis","Multimodal learning","Artificial Intelligence and Robotics","Biomedical Informatics","Radiology"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://commons.nmu.edu/theses/926","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Vinay Shashidhar"]},{"key":"dc:creator","label":"Author","values":["Noel-Rickert, Adele J"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.available","label":"Dc Date Available","values":["2026-03-27T07:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Math and Computer Science"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Science"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Pulmonary fibrosis","Chest X-ray","Multimodal deep learning","Medical imaging","Vision-language models","Radiology reports","Deep learning","Medical image analysis","Computer-aided diagnosis","Multimodal learning","Artificial Intelligence and Robotics","Biomedical Informatics","Radiology"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://commons.nmu.edu/theses/926"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Pulmonary fibrosis is a progressive interstitial lung disease characterized by the accumulation of fibrotic tissue within the lungs, leading to impaired respiratory function and reduced quality of life. Early detection is important for disease management; however, accurate diagnosis often relies on high-resolution computed tomography (CT), which may not be accessible in all clinical settings. Chest radiography provides a lower-cost and widely available imaging modality, but interpretation of chest X-rays for fibrotic disease can be challenging due to subtle radiographic patterns and overlapping anatomical structures. This thesis investigates the use of multimodal deep learning techniques to assist in pul- monary fibrosis detection by integrating chest X-ray images with corresponding radiology reports. Pretrained medical vision–language models, including BioViL-T, PubMedCLIP, and KAD, are used to generate embedding representations for both imaging and textual modalities. These embeddings are combined through a multimodal feature representation and processed by a neural network classifier to predict the presence of pulmonary fibro- sis. The proposed framework is evaluated using the PadChest dataset, which contains chest radiographs paired with radiology reports. Experimental results demonstrate that incorporating textual radiology reports alongside imaging features can improve diagnostic performance compared to models that rely solely on image-based representations. The findings highlight the potential of multimodal learning frameworks to better capture clinically relevant information by integrating visual imaging patterns with expert textual interpretation. This work contributes to the development of computational tools designed to support clinicians in pulmonary fibrosis assessment using widely available chest radiography data.
"]},{"key":"dc:title","label":"Title","values":["Deep Learning Based Approaches For Low Cost Defense Detection"]}]}],"canonical_facts":{"dc:contributor":["Vinay Shashidhar"],"dc:creator":["Noel-Rickert, Adele J"],"dc:date.available":["2026-03-27T07:00:00Z"],"dc:description.abstract":["Pulmonary fibrosis is a progressive interstitial lung disease characterized by the accumulation of fibrotic tissue within the lungs, leading to impaired respiratory function and reduced quality of life. Early detection is important for disease management; however, accurate diagnosis often relies on high-resolution computed tomography (CT), which may not be accessible in all clinical settings. Chest radiography provides a lower-cost and widely available imaging modality, but interpretation of chest X-rays for fibrotic disease can be challenging due to subtle radiographic patterns and overlapping anatomical structures. This thesis investigates the use of multimodal deep learning techniques to assist in pul- monary fibrosis detection by integrating chest X-ray images with corresponding radiology reports. Pretrained medical vision–language models, including BioViL-T, PubMedCLIP, and KAD, are used to generate embedding representations for both imaging and textual modalities. These embeddings are combined through a multimodal feature representation and processed by a neural network classifier to predict the presence of pulmonary fibro- sis. The proposed framework is evaluated using the PadChest dataset, which contains chest radiographs paired with radiology reports. Experimental results demonstrate that incorporating textual radiology reports alongside imaging features can improve diagnostic performance compared to models that rely solely on image-based representations. The findings highlight the potential of multimodal learning frameworks to better capture clinically relevant information by integrating visual imaging patterns with expert textual interpretation. This work contributes to the development of computational tools designed to support clinicians in pulmonary fibrosis assessment using widely available chest radiography data.
"],"dc:identifier":["https://commons.nmu.edu/theses/926"],"dc:subject":["Pulmonary fibrosis","Chest X-ray","Multimodal deep learning","Medical imaging","Vision-language models","Radiology reports","Deep learning","Medical image analysis","Computer-aided diagnosis","Multimodal learning","Artificial Intelligence and Robotics","Biomedical Informatics","Radiology"],"dc:title":["Deep Learning Based Approaches For Low Cost Defense Detection"],"thesis:degree_discipline":["Math and Computer Science"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["Master of Science"]},"updated_at":"2026-07-24T03:24:38Z"}