Related Experiment Video
Updated: Nov 19, 2025
07:15
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
7.2K
How to evaluate deep learning for cancer diagnostics - factors and recommendations
Roxana Daneshjou1, Bryan He2, David Ouyang3
1Department of Dermatology, Stanford University School of Medicine, Redwood City, CA, USA; Department of Biomedical Data Science, Stanford University, Stanford, CA, USA.
Biochimica Et Biophysica Acta. Reviews on Cancer
|January 29, 2021
Summary
Deep learning algorithms show promise for cancer diagnosis by analyzing large datasets, including clinical, radiological, and pathological images. This review explores current applications and outlines a future roadmap for integrating AI in cancer detection.
Area of Science:
- Artificial Intelligence in Oncology
- Medical Imaging Analysis
- Computational Pathology
Background:
- Vast amounts of data in cancer diagnosis offer opportunities for machine learning.
- Deep learning models improve predictive accuracy with increased data volume.
- Classifying samples (e.g., images, biomarkers) into categories like benign or cancerous is a key goal.
Purpose of the Study:
- To examine deep learning applications in cancer diagnosis using diverse data types.
- To present a systematic framework for evaluating clinical deep learning algorithms.
- To discuss future potential and provide a roadmap for AI implementation in oncology.
Main Methods:
- Review of deep learning implementations for cancer diagnosis.
- Analysis of clinical, radiological, and pathological image data.
- Systematic evaluation approach for clinical AI development.
Main Results:
- Deep learning effectively analyzes complex medical data for cancer prediction.
- Examples demonstrate successful application across various imaging modalities.
- A structured evaluation method is proposed for clinical AI.
Conclusions:
- Deep learning holds significant potential to advance cancer diagnosis.
- Future implementations require a strategic roadmap for integration.
- AI-driven insights can enhance predictive accuracy and clinical decision-making.
