AI Tool Improves Merkel Cell Carcinoma Prognosis

Merkel cell carcinoma is a rare but highly aggressive skin cancer. It is difficult to treat typically affecting older adults with weakened immune systems. It is difficult to treat, often diagnosed at advanced stages and linked to poor survival rate. A new hybrid machine-learning model is improving survival prediction for MCC. published in npj Digital Medicine the research demonstrates how artificial intelligence (AI) can outperform traditional cancer staging systems in providing personalized prognosis(1^✔ ✔Trusted Source
A hybrid machine learning approach for the personalized prognostication of aggressive skin cancers

Go to source).

MCC prognostication relies on tumor size and spread but this method often underestimates survival by ignoring other key factors such as patient demographics and socioeconomic status.

To address this, an international team of researchers combined machine learning with clinical expertise to develop DeepMerkel, a web-based AI tool that predicts treatment outcomes based on personal and tumor-specific features. The researchers suggest that this system could also be applied to other aggressive skin cancers improving clinical decision-making and giving patients better treatment choices.

DeepMerkel integrates deep learning and XGBoost a machine learning technique to provide personalized, time-dependent survival prediction. It is based on a broader set of factors like age, ethnicity, income level and tumor characteristics.

It demonstrates higher accuracy than traditional tumor staging systems. The model distinguished survival probabilities based on individual patient features, emphasizing the potential of AI-driven precision medicine.

They also developed a real-time web-based survival calculator tool that helps clinicians predict individual survival rates. By tailoring prognosis to each patient, this innovation could lead to better treatment planning and improved patient outcomes.

Advertisement

This highlights the transformative power of machine learning in oncology, setting a precedent for AI-driven prognostic models in other aggressive cancers. While further clinical validation is needed, DeepMerkel represents a major step forward in personalized medicine for MCC patients worldwide.

Reference:

1. A hybrid machine learning approach for the personalized prognostication of aggressive skin cancers - (https://www.nature.com/articles/s41746-024-01329-9)

Source-Medindia