Performance evaluation using MV instead of CWE.
<div><p>Skin lesions, including various abnormalities and potentially fatal skin cancers, require early detection for effective treatment. However, current methods often struggle to identify the precise areas responsible for these abnormalities after model dominance dispersion. To addres...
Saved in:
| Main Author: | |
|---|---|
| Published: |
2025
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1852020225363410944 |
|---|---|
| author | Anwar Hossain Efat (19942283) |
| author_facet | Anwar Hossain Efat (19942283) |
| author_role | author |
| dc.creator.none.fl_str_mv | Anwar Hossain Efat (19942283) |
| dc.date.none.fl_str_mv | 2025-05-20T17:25:40Z |
| dc.identifier.none.fl_str_mv | 10.1371/journal.pone.0321803.t022 |
| dc.relation.none.fl_str_mv | https://figshare.com/articles/dataset/Performance_evaluation_using_MV_instead_of_CWE_/29110807 |
| dc.rights.none.fl_str_mv | CC BY 4.0 info:eu-repo/semantics/openAccess |
| dc.subject.none.fl_str_mv | Medicine Cancer Space Science Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Information Systems not elsewhere classified require early detection precise areas responsible optimized regnet synergy optimal weight allocation offering valuable insights highlight critical regions ham1000 dataset demonstrates addressing challenges related model dominance dispersion layer ensemble approach enhance model interpretability skin lesion detection novel transfer learning div >< p layer weighted ensemble including various abnormalities cwe approach achieves 2 </ sup weighted ensemble >< sup model predictions skin lesions novel framework chi </ study fills practical applications impressive accuracy effective treatment current research boosts accuracy based framework art methods |
| dc.title.none.fl_str_mv | Performance evaluation using MV instead of CWE. |
| dc.type.none.fl_str_mv | Dataset info:eu-repo/semantics/publishedVersion dataset |
| description | <div><p>Skin lesions, including various abnormalities and potentially fatal skin cancers, require early detection for effective treatment. However, current methods often struggle to identify the precise areas responsible for these abnormalities after model dominance dispersion. To address this, we propose a novel Transfer Learning-based framework that integrates Optimized RegNet Synergy architectures and Attention-Triplet mechanisms—comprising channel attention, squeeze-excitation attention, and soft attention—combined with an advanced Ensemble Learning strategy. A significant gap in current research is the lack of techniques for optimal weight allocation in model predictions. Our study fills this gap by introducing the Weighted Ensemble (CWE) method, which is further enhanced into a Multi-Layer Weighted Ensemble (ML-CWE) to improve model aggregation across multiple layers. Evaluation on the HAM1000 dataset demonstrates that our ML-CWE approach achieves an impressive accuracy of 94.08%, outperforming existing state-of-the-art methods. To enhance model interpretability, we employ Gradient Class Activation Maps (Grad-CAM) to highlight critical regions of interest, improving both transparency and reliability. This work not only boosts accuracy but also facilitates early diagnosis, addressing challenges related to time, accessibility, and cost in skin lesion detection, and offering valuable insights for practical applications in dermatology.</p></div> |
| eu_rights_str_mv | openAccess |
| id | Manara_aaecdfd1c7004db9c2c2e217aaf88e63 |
| identifier_str_mv | 10.1371/journal.pone.0321803.t022 |
| network_acronym_str | Manara |
| network_name_str | ManaraRepo |
| oai_identifier_str | oai:figshare.com:article/29110807 |
| publishDate | 2025 |
| repository.mail.fl_str_mv | |
| repository.name.fl_str_mv | |
| repository_id_str | |
| rights_invalid_str_mv | CC BY 4.0 |
| spelling | Performance evaluation using MV instead of CWE.Anwar Hossain Efat (19942283)MedicineCancerSpace ScienceEnvironmental Sciences not elsewhere classifiedBiological Sciences not elsewhere classifiedMathematical Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedrequire early detectionprecise areas responsibleoptimized regnet synergyoptimal weight allocationoffering valuable insightshighlight critical regionsham1000 dataset demonstratesaddressing challenges relatedmodel dominance dispersionlayer ensemble approachenhance model interpretabilityskin lesion detectionnovel transfer learningdiv >< player weighted ensembleincluding various abnormalitiescwe approach achieves2 </ supweighted ensemble>< supmodel predictionsskin lesionsnovel frameworkchi </study fillspractical applicationsimpressive accuracyeffective treatmentcurrent researchboosts accuracybased frameworkart methods<div><p>Skin lesions, including various abnormalities and potentially fatal skin cancers, require early detection for effective treatment. However, current methods often struggle to identify the precise areas responsible for these abnormalities after model dominance dispersion. To address this, we propose a novel Transfer Learning-based framework that integrates Optimized RegNet Synergy architectures and Attention-Triplet mechanisms—comprising channel attention, squeeze-excitation attention, and soft attention—combined with an advanced Ensemble Learning strategy. A significant gap in current research is the lack of techniques for optimal weight allocation in model predictions. Our study fills this gap by introducing the Weighted Ensemble (CWE) method, which is further enhanced into a Multi-Layer Weighted Ensemble (ML-CWE) to improve model aggregation across multiple layers. Evaluation on the HAM1000 dataset demonstrates that our ML-CWE approach achieves an impressive accuracy of 94.08%, outperforming existing state-of-the-art methods. To enhance model interpretability, we employ Gradient Class Activation Maps (Grad-CAM) to highlight critical regions of interest, improving both transparency and reliability. This work not only boosts accuracy but also facilitates early diagnosis, addressing challenges related to time, accessibility, and cost in skin lesion detection, and offering valuable insights for practical applications in dermatology.</p></div>2025-05-20T17:25:40ZDatasetinfo:eu-repo/semantics/publishedVersiondataset10.1371/journal.pone.0321803.t022https://figshare.com/articles/dataset/Performance_evaluation_using_MV_instead_of_CWE_/29110807CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/291108072025-05-20T17:25:40Z |
| spellingShingle | Performance evaluation using MV instead of CWE. Anwar Hossain Efat (19942283) Medicine Cancer Space Science Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Information Systems not elsewhere classified require early detection precise areas responsible optimized regnet synergy optimal weight allocation offering valuable insights highlight critical regions ham1000 dataset demonstrates addressing challenges related model dominance dispersion layer ensemble approach enhance model interpretability skin lesion detection novel transfer learning div >< p layer weighted ensemble including various abnormalities cwe approach achieves 2 </ sup weighted ensemble >< sup model predictions skin lesions novel framework chi </ study fills practical applications impressive accuracy effective treatment current research boosts accuracy based framework art methods |
| status_str | publishedVersion |
| title | Performance evaluation using MV instead of CWE. |
| title_full | Performance evaluation using MV instead of CWE. |
| title_fullStr | Performance evaluation using MV instead of CWE. |
| title_full_unstemmed | Performance evaluation using MV instead of CWE. |
| title_short | Performance evaluation using MV instead of CWE. |
| title_sort | Performance evaluation using MV instead of CWE. |
| topic | Medicine Cancer Space Science Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Information Systems not elsewhere classified require early detection precise areas responsible optimized regnet synergy optimal weight allocation offering valuable insights highlight critical regions ham1000 dataset demonstrates addressing challenges related model dominance dispersion layer ensemble approach enhance model interpretability skin lesion detection novel transfer learning div >< p layer weighted ensemble including various abnormalities cwe approach achieves 2 </ sup weighted ensemble >< sup model predictions skin lesions novel framework chi </ study fills practical applications impressive accuracy effective treatment current research boosts accuracy based framework art methods |