Background: Skin cancer is among the most prevalent malignancies globally. Early detection remains the most crucial component in improving patient outcomes. However, current diagnostic pathways depend on dermoscopy and invasive biopsies. These methods are sometimes costly or distressing for the patients undergoing them. Hydrogel-based wearable patches have emerged as a promising class of biosensors that can interface directly with skin tissue and interstitial fluid to capture disease-related biomarkers including inflammatory cytokines, non-invasively. Despite this potential, the translation of raw signals into clinically meaningful diagnostic outputs remains a challenge. The integration of machine learning with hydrogel biosensor platforms offers a pathway toward automated, non-destructive screening for early-stage skin cancer biomarkers.
Aim: This study aims to propose a machine learning framework using hydrogel-based wearable skin patches for non-invasive detection and characterization of biomarkers associated with skin cancer.
Methods: This study proposes a conceptual framework where a flexible hydrogel patch is applied directly to the skin, enabling continuous sensing of biomarkers including inflammatory cytokines and tumor-associated proteins. The hydrogel matrix is designed to incorporate electrochemical or optical nanosensors. Output data is fed into a machine for automated analysis. Supervised classification models including logistic regression are proposed as baselines. Model performance will be evaluated using standard metrics such as AUC and sensitivity. All detection is performed non-invasively, with no skin disruption, aligning with core non-destructive testing principles.
Expected Results: Integration of hydrogel-based sensing with machine learning will enable accurate and early-stage detection of skin cancer-associated biomarker profiles, outperforming conventional approaches while maintaining a non-invasive detection technique.
Conclusion: This work presents a novel framework for non-invasive skin cancer biomarker detection, combining advanced hydrogel sensing technology with machine learning to support earlier diagnostic screening.
