Abstract :

Melanoma Ranks As One Of The Most Lethal Skin Cancers, With Metastasis Responsible For The Majority Of Associated Deaths. Early And Accurate Identification Of Metastasis Is Crucial For Informing Treatment Choices And Enhancing Patient Survival. Previous Deep Learning Methods Have Shown Excellent Results In Melanoma Classification But Frequently Experience Drawbacks Like Dependence On A Single Backbone, Limited Feature Representation, Or Failure To Integrate Diverse Data From Various Medical Sources. In This Study, We Introduce A Hybrid Deep Learning Model That Combines Dermoscopic Images From The ISIC Dataset With Clinical And Genomic Information From The TCGA-SKCM Cohort To Forecast Melanoma Metastasis. The Suggested Model Integrates The Unique Advantages Of ResNet For Hierarchical Feature Extraction And EfficientNet-B0 For Enhanced Multi-scale Representation, Allowing For A More Profound And Distinguishing Comprehension Of Lesion Morphology. A Fusion Module Combines Deep Visual Features With Patient-specific Characteristics To Improve Metastasis Forecasting. Experimental Assessment Verifies That Our Hybrid Architecture Provides Enhanced Accuracy, Robustness, And Generalization Over Single-model Baselines. The Findings Emphasize The Capability Of Combining Multi-source Data With Efficient Hybrid Networks To Enhance Accurate Metastasis Risk Evaluation And Clinical Decision-making In Managing Melanoma

Published:

03-4-1-2026

Issue:

Vol. 26 No. 4-1 (2026)

Page Nos:

1-8

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite