A Deep Learning Enabled Chest X-Ray Abnormality Detection Model for Radiology Assistance.
Interpreting chest X-ray images is a challenging task for radiologists due to the complexity of identifying abnormalities accurately. This difficulty persists even among experienced radiologists, leading to potential errors and delays in diagnosis (Pang et al., 2021). Traditional methods of interpre...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
Kabale University
2024
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/20.500.12493/2639 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823844183120543744 |
|---|---|
| author | Ssempeebwa, Phillip Ainembabazi, Patience |
| author_facet | Ssempeebwa, Phillip Ainembabazi, Patience |
| author_sort | Ssempeebwa, Phillip |
| collection | KAB-DR |
| description | Interpreting chest X-ray images is a challenging task for radiologists due to the complexity of identifying abnormalities accurately. This difficulty persists even among experienced radiologists, leading to potential errors and delays in diagnosis (Pang et al., 2021). Traditional methods of
interpretation rely heavily on manual inspection, which can be time-consuming and prone to human error. Therefore, there is a critical need for innovative solutions to improve the efficiency and accuracy of chest X-ray diagnosis. The main objective of this study is to develop and implement a
Deep Learning-Enabled Chest X-ray Abnormality Detection model with automated report generation. The aim is to transform the traditional diagnostic workflow by automating both the detection and reporting processes of chest abnormalities. By doing so, this research seeks to prevent
the burden on radiologists and enhance patient outcomes by providing timely and accurate diagnoses. The methodology employed in this research involves several key steps. Firstly, the model undergoes an image preprocessing phase, where chest X-ray images are standardized and
segmented into distinct anatomical regions, focusing on the upper, lower, and middle sections of the lungs. Simultaneously, dedicated binary-classification deep learning models are deployed to analyze each segmented region individually, effectively detecting specific abnormalities such as
cardiomegaly, lung effusion, and consolidation with high accuracy. The outputs from these models are then consolidated into a cohesive 'result code,' summarizing the presence or absence of abnormalities across all segmented regions. Finally, using this result code, the model generates a
comprehensive radiology report, integrating the detected abnormalities into an easily interpretable format for healthcare professionals. Through this methodology, this research is designed to improve the radiologist's workload and enhance the accuracy of chest X-ray diagnosis. By automating the
detection and reporting process, the researchers anticipate a reduction in diagnostic time, improved accuracy, and increased efficiency in the overall interpretation of chest X-ray images. This research contributes to the ongoing efforts to integrate deep learning into medical imaging workflows for
more effective and reliable diagnostic outcomes. |
| format | Thesis |
| id | oai:idr.kab.ac.ug:20.500.12493-2639 |
| institution | KAB-DR |
| language | English |
| publishDate | 2024 |
| publisher | Kabale University |
| record_format | dspace |
| spelling | oai:idr.kab.ac.ug:20.500.12493-26392024-12-31T00:01:14Z A Deep Learning Enabled Chest X-Ray Abnormality Detection Model for Radiology Assistance. Ssempeebwa, Phillip Ainembabazi, Patience Deep Learning Enabled Chest X-Ray Abnormality Detection Model Radiology Assistance Interpreting chest X-ray images is a challenging task for radiologists due to the complexity of identifying abnormalities accurately. This difficulty persists even among experienced radiologists, leading to potential errors and delays in diagnosis (Pang et al., 2021). Traditional methods of interpretation rely heavily on manual inspection, which can be time-consuming and prone to human error. Therefore, there is a critical need for innovative solutions to improve the efficiency and accuracy of chest X-ray diagnosis. The main objective of this study is to develop and implement a Deep Learning-Enabled Chest X-ray Abnormality Detection model with automated report generation. The aim is to transform the traditional diagnostic workflow by automating both the detection and reporting processes of chest abnormalities. By doing so, this research seeks to prevent the burden on radiologists and enhance patient outcomes by providing timely and accurate diagnoses. The methodology employed in this research involves several key steps. Firstly, the model undergoes an image preprocessing phase, where chest X-ray images are standardized and segmented into distinct anatomical regions, focusing on the upper, lower, and middle sections of the lungs. Simultaneously, dedicated binary-classification deep learning models are deployed to analyze each segmented region individually, effectively detecting specific abnormalities such as cardiomegaly, lung effusion, and consolidation with high accuracy. The outputs from these models are then consolidated into a cohesive 'result code,' summarizing the presence or absence of abnormalities across all segmented regions. Finally, using this result code, the model generates a comprehensive radiology report, integrating the detected abnormalities into an easily interpretable format for healthcare professionals. Through this methodology, this research is designed to improve the radiologist's workload and enhance the accuracy of chest X-ray diagnosis. By automating the detection and reporting process, the researchers anticipate a reduction in diagnostic time, improved accuracy, and increased efficiency in the overall interpretation of chest X-ray images. This research contributes to the ongoing efforts to integrate deep learning into medical imaging workflows for more effective and reliable diagnostic outcomes. 2024-12-30T12:17:09Z 2024-12-30T12:17:09Z 2024 Thesis Ssempeebwa, P., & Ainembabazi, P. (2024). A Deep Learning Enabled Chest X-Ray Abnormality Detection Model for Radiology Assistance. Kabale: Kabale University. http://hdl.handle.net/20.500.12493/2639 en Attribution-NonCommercial-NoDerivs 3.0 United States http://creativecommons.org/licenses/by-nc-nd/3.0/us/ application/pdf Kabale University |
| spellingShingle | Deep Learning Enabled Chest X-Ray Abnormality Detection Model Radiology Assistance Ssempeebwa, Phillip Ainembabazi, Patience A Deep Learning Enabled Chest X-Ray Abnormality Detection Model for Radiology Assistance. |
| title | A Deep Learning Enabled Chest X-Ray Abnormality Detection Model for Radiology Assistance. |
| title_full | A Deep Learning Enabled Chest X-Ray Abnormality Detection Model for Radiology Assistance. |
| title_fullStr | A Deep Learning Enabled Chest X-Ray Abnormality Detection Model for Radiology Assistance. |
| title_full_unstemmed | A Deep Learning Enabled Chest X-Ray Abnormality Detection Model for Radiology Assistance. |
| title_short | A Deep Learning Enabled Chest X-Ray Abnormality Detection Model for Radiology Assistance. |
| title_sort | deep learning enabled chest x ray abnormality detection model for radiology assistance |
| topic | Deep Learning Enabled Chest X-Ray Abnormality Detection Model Radiology Assistance |
| url | http://hdl.handle.net/20.500.12493/2639 |
| work_keys_str_mv | AT ssempeebwaphillip adeeplearningenabledchestxrayabnormalitydetectionmodelforradiologyassistance AT ainembabazipatience adeeplearningenabledchestxrayabnormalitydetectionmodelforradiologyassistance AT ssempeebwaphillip deeplearningenabledchestxrayabnormalitydetectionmodelforradiologyassistance AT ainembabazipatience deeplearningenabledchestxrayabnormalitydetectionmodelforradiologyassistance |