AI-Based Multiclass Grading of Hepatic Steatosis From B-Mode Ultrasound: Generalization Across Modalities and Clinical Comparison With Radiologists
<p dir="ltr">Non-alcoholic fatty liver disease (NAFLD) is a growing public health challenge, underscoring the need for scalable, non-invasive tools to grade hepatic steatosis. Although B-mode ultrasound is accessible and safe, its reliability is limited by operator and scanner variab...
محفوظ في:
| المؤلف الرئيسي: | |
|---|---|
| مؤلفون آخرون: | , , , , , , , , , , , , , , , |
| منشور في: |
2025
|
| الموضوعات: | |
| الوسوم: |
إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
|
| _version_ | 1864513524092370944 |
|---|---|
| author | Fahad Muflih Alshagathrh (18427950) |
| author2 | Haider Dhia Zubaydi (18519360) Mahmood Alzubaidi (15740693) Abdulaziz Alosaimi (11475559) Raneem Mohammed Al Saqer (23073700) Abdullah Mutlaq Alzahrani (23073703) Mei Khalid Alfaqiri (23073706) Mohamed Rajab Elzahrani (23073709) Khalid Alswat (13047418) Ali Aldhebaib (23073712) Bushra Alahmadi (23073715) Meteb Alkubeyyer (23073718) Amani Alsadoon (23073721) Maram Alkhamash (23073724) Jawad Ahmad Alraimi (23073727) Jens Schneider (16885948) Mowafa Househ (9154124) |
| author2_role | author author author author author author author author author author author author author author author author |
| author_facet | Fahad Muflih Alshagathrh (18427950) Haider Dhia Zubaydi (18519360) Mahmood Alzubaidi (15740693) Abdulaziz Alosaimi (11475559) Raneem Mohammed Al Saqer (23073700) Abdullah Mutlaq Alzahrani (23073703) Mei Khalid Alfaqiri (23073706) Mohamed Rajab Elzahrani (23073709) Khalid Alswat (13047418) Ali Aldhebaib (23073712) Bushra Alahmadi (23073715) Meteb Alkubeyyer (23073718) Amani Alsadoon (23073721) Maram Alkhamash (23073724) Jawad Ahmad Alraimi (23073727) Jens Schneider (16885948) Mowafa Househ (9154124) |
| author_role | author |
| dc.creator.none.fl_str_mv | Fahad Muflih Alshagathrh (18427950) Haider Dhia Zubaydi (18519360) Mahmood Alzubaidi (15740693) Abdulaziz Alosaimi (11475559) Raneem Mohammed Al Saqer (23073700) Abdullah Mutlaq Alzahrani (23073703) Mei Khalid Alfaqiri (23073706) Mohamed Rajab Elzahrani (23073709) Khalid Alswat (13047418) Ali Aldhebaib (23073712) Bushra Alahmadi (23073715) Meteb Alkubeyyer (23073718) Amani Alsadoon (23073721) Maram Alkhamash (23073724) Jawad Ahmad Alraimi (23073727) Jens Schneider (16885948) Mowafa Househ (9154124) |
| dc.date.none.fl_str_mv | 2025-10-21T12:00:00Z |
| dc.identifier.none.fl_str_mv | 10.1109/access.2025.3617778 |
| dc.relation.none.fl_str_mv | https://figshare.com/articles/journal_contribution/AI-Based_Multiclass_Grading_of_Hepatic_Steatosis_From_B-Mode_Ultrasound_Generalization_Across_Modalities_and_Clinical_Comparison_With_Radiologists/31169242 |
| dc.rights.none.fl_str_mv | CC BY 4.0 info:eu-repo/semantics/openAccess |
| dc.subject.none.fl_str_mv | Biomedical and clinical sciences Clinical sciences Engineering Biomedical engineering Health sciences Health services and systems Information and computing sciences Artificial intelligence Artificial intelligence deep learning ultrasound imaging hepatic steatosis multiclass classification domain adaptation diagnostic accuracy inter-rater reliability biopsy ground truth non-alcoholic fatty liver disease (NAFLD) Training Adaptation models Accuracy Urban areas Liver diseases Deep learning Benchmark testing |
| dc.title.none.fl_str_mv | AI-Based Multiclass Grading of Hepatic Steatosis From B-Mode Ultrasound: Generalization Across Modalities and Clinical Comparison With Radiologists |
| dc.type.none.fl_str_mv | Text Journal contribution info:eu-repo/semantics/publishedVersion text contribution to journal |
| description | <p dir="ltr">Non-alcoholic fatty liver disease (NAFLD) is a growing public health challenge, underscoring the need for scalable, non-invasive tools to grade hepatic steatosis. Although B-mode ultrasound is accessible and safe, its reliability is limited by operator and scanner variability. We present the Deep Domain Adaptation Neural Network (DDANN), a deep learning system for multiclass steatosis classification (Normal, Mild, Moderate, Severe) from ultrasound that emphasizes cross-device generalizability. To mitigate distribution shifts across scanners (LOGIQ, iU22, EPIQ), DDANN combines a MobileNetV2 backbone with triplet loss, entropy-based domain adaptation, and preprocessing that includes speckle suppression, percentile normalization, and LOGIQ-specific harmonization. Trained on a biopsy-confirmed, multi-institutional cohort (primarily LOGIQ and iU22), the model was externally validated on an unseen EPIQ test set of 1,083 images from 47 patients, achieving 98.71% accuracy, 0.9872 macro <i>F</i><sub><em>1</em></sub> -score, and 0.9998 AUC-ROC, outperforming baselines. In a separate radiologist–AI comparison on 224 biopsy-confirmed images not used for training or validation, the AI reached 91.96% accuracy, significantly exceeding radiologists’ 19.64%–31.70% (McNemar’s test, <i>p</i><0.001 ), with strong agreement to ground truth (<i>κ</i>=0.893) versus radiologists’ poor-to-slight agreement (<i>κ</i>=0.006 –0.194). The AI maintained balanced class-wise <i>F</i><sub><em>1</em></sub> -scores (0.90–0.94), while radiologists struggled, particularly with Mild and Moderate cases, and exhibited substantial inter-reader variability (<i>κ</i>=0.068 –0.648). These results demonstrate robust cross-device performance and support integrating AI as a reliable second reader or primary screening tool to reduce subjectivity in steatosis assessment.</p><h2 dir="ltr">Other Information</h2><p dir="ltr">Published in: IEEE Access<br>License: <a href="https://creativecommons.org/licenses/by/4.0/deed.en" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1109/access.2025.3617778" target="_blank">https://dx.doi.org/10.1109/access.2025.3617778</a></p> |
| eu_rights_str_mv | openAccess |
| id | Manara2_93f4ec7bb421afe2754d39356a4fe9ed |
| identifier_str_mv | 10.1109/access.2025.3617778 |
| network_acronym_str | Manara2 |
| network_name_str | Manara2 |
| oai_identifier_str | oai:figshare.com:article/31169242 |
| publishDate | 2025 |
| repository.mail.fl_str_mv | |
| repository.name.fl_str_mv | |
| repository_id_str | |
| rights_invalid_str_mv | CC BY 4.0 |
| spelling | AI-Based Multiclass Grading of Hepatic Steatosis From B-Mode Ultrasound: Generalization Across Modalities and Clinical Comparison With RadiologistsFahad Muflih Alshagathrh (18427950)Haider Dhia Zubaydi (18519360)Mahmood Alzubaidi (15740693)Abdulaziz Alosaimi (11475559)Raneem Mohammed Al Saqer (23073700)Abdullah Mutlaq Alzahrani (23073703)Mei Khalid Alfaqiri (23073706)Mohamed Rajab Elzahrani (23073709)Khalid Alswat (13047418)Ali Aldhebaib (23073712)Bushra Alahmadi (23073715)Meteb Alkubeyyer (23073718)Amani Alsadoon (23073721)Maram Alkhamash (23073724)Jawad Ahmad Alraimi (23073727)Jens Schneider (16885948)Mowafa Househ (9154124)Biomedical and clinical sciencesClinical sciencesEngineeringBiomedical engineeringHealth sciencesHealth services and systemsInformation and computing sciencesArtificial intelligenceArtificial intelligencedeep learningultrasound imaginghepatic steatosismulticlass classificationdomain adaptationdiagnostic accuracyinter-rater reliabilitybiopsy ground truthnon-alcoholic fatty liver disease (NAFLD)TrainingAdaptation modelsAccuracyUrban areasLiver diseasesDeep learningBenchmark testing<p dir="ltr">Non-alcoholic fatty liver disease (NAFLD) is a growing public health challenge, underscoring the need for scalable, non-invasive tools to grade hepatic steatosis. Although B-mode ultrasound is accessible and safe, its reliability is limited by operator and scanner variability. We present the Deep Domain Adaptation Neural Network (DDANN), a deep learning system for multiclass steatosis classification (Normal, Mild, Moderate, Severe) from ultrasound that emphasizes cross-device generalizability. To mitigate distribution shifts across scanners (LOGIQ, iU22, EPIQ), DDANN combines a MobileNetV2 backbone with triplet loss, entropy-based domain adaptation, and preprocessing that includes speckle suppression, percentile normalization, and LOGIQ-specific harmonization. Trained on a biopsy-confirmed, multi-institutional cohort (primarily LOGIQ and iU22), the model was externally validated on an unseen EPIQ test set of 1,083 images from 47 patients, achieving 98.71% accuracy, 0.9872 macro <i>F</i><sub><em>1</em></sub> -score, and 0.9998 AUC-ROC, outperforming baselines. In a separate radiologist–AI comparison on 224 biopsy-confirmed images not used for training or validation, the AI reached 91.96% accuracy, significantly exceeding radiologists’ 19.64%–31.70% (McNemar’s test, <i>p</i><0.001 ), with strong agreement to ground truth (<i>κ</i>=0.893) versus radiologists’ poor-to-slight agreement (<i>κ</i>=0.006 –0.194). The AI maintained balanced class-wise <i>F</i><sub><em>1</em></sub> -scores (0.90–0.94), while radiologists struggled, particularly with Mild and Moderate cases, and exhibited substantial inter-reader variability (<i>κ</i>=0.068 –0.648). These results demonstrate robust cross-device performance and support integrating AI as a reliable second reader or primary screening tool to reduce subjectivity in steatosis assessment.</p><h2 dir="ltr">Other Information</h2><p dir="ltr">Published in: IEEE Access<br>License: <a href="https://creativecommons.org/licenses/by/4.0/deed.en" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1109/access.2025.3617778" target="_blank">https://dx.doi.org/10.1109/access.2025.3617778</a></p>2025-10-21T12:00:00ZTextJournal contributioninfo:eu-repo/semantics/publishedVersiontextcontribution to journal10.1109/access.2025.3617778https://figshare.com/articles/journal_contribution/AI-Based_Multiclass_Grading_of_Hepatic_Steatosis_From_B-Mode_Ultrasound_Generalization_Across_Modalities_and_Clinical_Comparison_With_Radiologists/31169242CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/311692422025-10-21T12:00:00Z |
| spellingShingle | AI-Based Multiclass Grading of Hepatic Steatosis From B-Mode Ultrasound: Generalization Across Modalities and Clinical Comparison With Radiologists Fahad Muflih Alshagathrh (18427950) Biomedical and clinical sciences Clinical sciences Engineering Biomedical engineering Health sciences Health services and systems Information and computing sciences Artificial intelligence Artificial intelligence deep learning ultrasound imaging hepatic steatosis multiclass classification domain adaptation diagnostic accuracy inter-rater reliability biopsy ground truth non-alcoholic fatty liver disease (NAFLD) Training Adaptation models Accuracy Urban areas Liver diseases Deep learning Benchmark testing |
| status_str | publishedVersion |
| title | AI-Based Multiclass Grading of Hepatic Steatosis From B-Mode Ultrasound: Generalization Across Modalities and Clinical Comparison With Radiologists |
| title_full | AI-Based Multiclass Grading of Hepatic Steatosis From B-Mode Ultrasound: Generalization Across Modalities and Clinical Comparison With Radiologists |
| title_fullStr | AI-Based Multiclass Grading of Hepatic Steatosis From B-Mode Ultrasound: Generalization Across Modalities and Clinical Comparison With Radiologists |
| title_full_unstemmed | AI-Based Multiclass Grading of Hepatic Steatosis From B-Mode Ultrasound: Generalization Across Modalities and Clinical Comparison With Radiologists |
| title_short | AI-Based Multiclass Grading of Hepatic Steatosis From B-Mode Ultrasound: Generalization Across Modalities and Clinical Comparison With Radiologists |
| title_sort | AI-Based Multiclass Grading of Hepatic Steatosis From B-Mode Ultrasound: Generalization Across Modalities and Clinical Comparison With Radiologists |
| topic | Biomedical and clinical sciences Clinical sciences Engineering Biomedical engineering Health sciences Health services and systems Information and computing sciences Artificial intelligence Artificial intelligence deep learning ultrasound imaging hepatic steatosis multiclass classification domain adaptation diagnostic accuracy inter-rater reliability biopsy ground truth non-alcoholic fatty liver disease (NAFLD) Training Adaptation models Accuracy Urban areas Liver diseases Deep learning Benchmark testing |