Adaptive Evolutionary Optimization of Deep Learning Architectures for Focused Liver Ultrasound Image Segmentation
<b>Background:</b> Liver ultrasound segmentation is challenging due to low image quality and variability. While deep learning (DL) models have been widely applied for medical segmentation, generic pre-configured models may not meet the specific requirements for targeted areas in liver ul...
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/15/2/117 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832588739501621248 |
|---|---|
| author | Ali Zifan Katelyn Zhao Madilyn Lee Zihan Peng Laura J. Roney Sarayu Pai Jake T. Weeks Michael S. Middleton Ahmed El Kaffas Jeffrey B. Schwimmer Claude B. Sirlin |
| author_facet | Ali Zifan Katelyn Zhao Madilyn Lee Zihan Peng Laura J. Roney Sarayu Pai Jake T. Weeks Michael S. Middleton Ahmed El Kaffas Jeffrey B. Schwimmer Claude B. Sirlin |
| author_sort | Ali Zifan |
| collection | DOAJ |
| description | <b>Background:</b> Liver ultrasound segmentation is challenging due to low image quality and variability. While deep learning (DL) models have been widely applied for medical segmentation, generic pre-configured models may not meet the specific requirements for targeted areas in liver ultrasound. Quantitative ultrasound (QUS) is emerging as a promising tool for liver fat measurement; however, accurately segmenting regions of interest within liver ultrasound images remains a challenge. <b>Methods:</b> We introduce a generalizable framework using an adaptive evolutionary genetic algorithm to optimize deep learning models, specifically U-Net, for focused liver segmentation. The algorithm simultaneously adjusts the depth (number of layers) and width (neurons per layer) of the network, dropout, and skip connections. Various architecture configurations are evaluated based on segmentation performance to find the optimal model for liver ultrasound images. <b>Results:</b> The model with a depth of 4 and filter sizes of [16, 64, 128, 256] achieved the highest mean adjusted Dice score of 0.921, outperforming the other configurations, using three-fold cross-validation with early stoppage. <b>Conclusions:</b> Adaptive evolutionary optimization enhances the deep learning architecture for liver ultrasound segmentation. Future work may extend this optimization to other imaging modalities and deep learning architectures. |
| format | Article |
| id | doaj-art-fe100a9e8ec34b35b1eaea1b5f635530 |
| institution | Kabale University |
| issn | 2075-4418 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj-art-fe100a9e8ec34b35b1eaea1b5f6355302025-01-24T13:28:47ZengMDPI AGDiagnostics2075-44182025-01-0115211710.3390/diagnostics15020117Adaptive Evolutionary Optimization of Deep Learning Architectures for Focused Liver Ultrasound Image SegmentationAli Zifan0Katelyn Zhao1Madilyn Lee2Zihan Peng3Laura J. Roney4Sarayu Pai5Jake T. Weeks6Michael S. Middleton7Ahmed El Kaffas8Jeffrey B. Schwimmer9Claude B. Sirlin10Division of Gastroenterology and Hepatology, University of California San Diego, San Diego, CA 92093, USADivision of Gastroenterology and Hepatology, University of California San Diego, San Diego, CA 92093, USADivision of Gastroenterology and Hepatology, University of California San Diego, San Diego, CA 92093, USADivision of Gastroenterology and Hepatology, University of California San Diego, San Diego, CA 92093, USADivision of Gastroenterology and Hepatology, University of California San Diego, San Diego, CA 92093, USADivision of Gastroenterology and Hepatology, University of California San Diego, San Diego, CA 92093, USALiver Imaging Group, Department of Radiology, University of California San Diego, San Diego, CA 92093, USALiver Imaging Group, Department of Radiology, University of California San Diego, San Diego, CA 92093, USALiver Imaging Group, Department of Radiology, University of California San Diego, San Diego, CA 92093, USADepartment of Pediatrics, Division of Gastroenterology, Hepatology, and Nutrition, University of California San Diego School of Medicine, La Jolla, CA 92093, USALiver Imaging Group, Department of Radiology, University of California San Diego, San Diego, CA 92093, USA<b>Background:</b> Liver ultrasound segmentation is challenging due to low image quality and variability. While deep learning (DL) models have been widely applied for medical segmentation, generic pre-configured models may not meet the specific requirements for targeted areas in liver ultrasound. Quantitative ultrasound (QUS) is emerging as a promising tool for liver fat measurement; however, accurately segmenting regions of interest within liver ultrasound images remains a challenge. <b>Methods:</b> We introduce a generalizable framework using an adaptive evolutionary genetic algorithm to optimize deep learning models, specifically U-Net, for focused liver segmentation. The algorithm simultaneously adjusts the depth (number of layers) and width (neurons per layer) of the network, dropout, and skip connections. Various architecture configurations are evaluated based on segmentation performance to find the optimal model for liver ultrasound images. <b>Results:</b> The model with a depth of 4 and filter sizes of [16, 64, 128, 256] achieved the highest mean adjusted Dice score of 0.921, outperforming the other configurations, using three-fold cross-validation with early stoppage. <b>Conclusions:</b> Adaptive evolutionary optimization enhances the deep learning architecture for liver ultrasound segmentation. Future work may extend this optimization to other imaging modalities and deep learning architectures.https://www.mdpi.com/2075-4418/15/2/117ultrasound liver segmentationdeep learning optimizationevolutionary genetic algorithm |
| spellingShingle | Ali Zifan Katelyn Zhao Madilyn Lee Zihan Peng Laura J. Roney Sarayu Pai Jake T. Weeks Michael S. Middleton Ahmed El Kaffas Jeffrey B. Schwimmer Claude B. Sirlin Adaptive Evolutionary Optimization of Deep Learning Architectures for Focused Liver Ultrasound Image Segmentation Diagnostics ultrasound liver segmentation deep learning optimization evolutionary genetic algorithm |
| title | Adaptive Evolutionary Optimization of Deep Learning Architectures for Focused Liver Ultrasound Image Segmentation |
| title_full | Adaptive Evolutionary Optimization of Deep Learning Architectures for Focused Liver Ultrasound Image Segmentation |
| title_fullStr | Adaptive Evolutionary Optimization of Deep Learning Architectures for Focused Liver Ultrasound Image Segmentation |
| title_full_unstemmed | Adaptive Evolutionary Optimization of Deep Learning Architectures for Focused Liver Ultrasound Image Segmentation |
| title_short | Adaptive Evolutionary Optimization of Deep Learning Architectures for Focused Liver Ultrasound Image Segmentation |
| title_sort | adaptive evolutionary optimization of deep learning architectures for focused liver ultrasound image segmentation |
| topic | ultrasound liver segmentation deep learning optimization evolutionary genetic algorithm |
| url | https://www.mdpi.com/2075-4418/15/2/117 |
| work_keys_str_mv | AT alizifan adaptiveevolutionaryoptimizationofdeeplearningarchitecturesforfocusedliverultrasoundimagesegmentation AT katelynzhao adaptiveevolutionaryoptimizationofdeeplearningarchitecturesforfocusedliverultrasoundimagesegmentation AT madilynlee adaptiveevolutionaryoptimizationofdeeplearningarchitecturesforfocusedliverultrasoundimagesegmentation AT zihanpeng adaptiveevolutionaryoptimizationofdeeplearningarchitecturesforfocusedliverultrasoundimagesegmentation AT laurajroney adaptiveevolutionaryoptimizationofdeeplearningarchitecturesforfocusedliverultrasoundimagesegmentation AT sarayupai adaptiveevolutionaryoptimizationofdeeplearningarchitecturesforfocusedliverultrasoundimagesegmentation AT jaketweeks adaptiveevolutionaryoptimizationofdeeplearningarchitecturesforfocusedliverultrasoundimagesegmentation AT michaelsmiddleton adaptiveevolutionaryoptimizationofdeeplearningarchitecturesforfocusedliverultrasoundimagesegmentation AT ahmedelkaffas adaptiveevolutionaryoptimizationofdeeplearningarchitecturesforfocusedliverultrasoundimagesegmentation AT jeffreybschwimmer adaptiveevolutionaryoptimizationofdeeplearningarchitecturesforfocusedliverultrasoundimagesegmentation AT claudebsirlin adaptiveevolutionaryoptimizationofdeeplearningarchitecturesforfocusedliverultrasoundimagesegmentation |