Authors: Al Amin, Kamrul Hasan, Saleh Zein-Sabatto, Deo Chimba, Imtiaz Ahmed, Tariqul Islam
Published on: March 07, 2024
Impact Score: 8.0
Arxiv code: Arxiv:2403.04130
Summary
- What is new: Leveraging a custom XAI framework with techniques like LIME, SHAP, and Grad-Cam in the AIoMT domain for improving healthcare systems, particularly in transparent and interpretative decision-making in medical applications.
- Why this is important: The complexity of AI models in healthcare and the crucial need for transparent, interpretable decision-making.
- What the research proposes: A custom Explainable Artificial Intelligence (XAI) framework designed for AIoMT, utilizing ensemble-based DL methodologies for accurate and trustworthy medical diagnoses.
- Results: High precision, recall, and F1 scores with 99% training accuracy and 98% validation accuracy in brain tumor detection, demonstrating the framework’s effectiveness in making precise and reliable diagnoses.
Technical Details
Technological frameworks used: Custom XAI
Models used: LIME, SHAP, Grad-Cam, multiple CNNs with a majority voting technique
Data used: AIoMT for brain tumor detection
Potential Impact
Healthcare industry, particularly companies and markets involved in medical diagnostics and AI solutions for healthcare.
Want to implement this idea in a business?
We have generated a startup concept here: ClearDiagnose.
Leave a Reply