Here I will be summarizing worldwide papers about the use of AI in Healthcare and medicine. On each paper, it will be covered some main characteristics (discussed later). This project aims to help review and help about the "reproductibility crisis" around AI.
As a first goal, I want to cover the articles (that, of course, are applicable) in the reference section of "High-performance medicine: the convergence of human and artificial intelligence". Naturally, this goal can be expanded/renewed in order to cover more recent research on the field.
Contributions are welcomed.
This is what will be shown or analyzed:
- Title
- Date
- Medical specialty
- Cardiology, mental health,
- Target condition
- Diabetes, breast tumor, lung cancer, ...
- Algorithm used
- CNN, Random Forest, k-NN, ...
- Hyperparameters were given?
- Yes, no
- Code availability
- Yes, no
- Dataset name
- Pima Indians Diabetes, ChestX-ray14, ...
- Dataset availability
- Yes, no
- Data type (image, audio, numeric, nominal, ...)
- Tabular, text, Image 2D, Image 3D, ...
- Metrics used to measure the performance
- AUC, accuracy, sensibility, sensitivity, ...
- Explainability
- Heatmap, ...
- ChestX-ray8: Hospital-scale Chest X-ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases
- Thoracic Disease Identification and Localization with Limited Supervision
- Deep learning in chest radiography: Detection of findings and presence of change
- Clinically applicable deep learning for diagnosis and referral in retinal disease
- Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network
- The Artificial Intelligence Clinician learns optimal treatment strategies for sepsis in intensive care