- week01:
- Big Data in Healthcare
- EHR System in the UK and USA
- MIMIC Critical Care Dataset: The Impact
- Data Usage Requirements
- week02:
- MIMIC-III Data Linkage
- MIMIC-III as a Relational Database
- MIMIC-III - Descriptive Statistics
- Vital Signs Extraction for a Single Patient
- week03
- Introduction to International Classification of Disease System
- Evolution of the ICD System
- ICD-9 and MIMIC-III
- From ICD-9 to ICD-10 (ICD-11)
- week04
- Concepts in MIMIC III
- Relation of Catheterization to Mortality: An Example Study
- Data Extraction
- week01:
- Deep Learning and Artificial Intelligence
- Multi-Layer Perceptron
- Optimization of a Multi-Layer Perceptron (Part 1)
- Optimization of a Mutli-Layer Perceptrion (Part 2)
- Preprocessing of ECG Signal
- week02:
- Validation of Machine Learning Models
- Convolutional Neural Networks
- Recurrent Neural Networks
- week03
- Benchmark Deep Learning Models with EHR - Part 1
- Benchmark Deep Learning Models with EHR - Part 2
- Imputation Strategies
- Deep Learning Imputation Strategies
- week04
- Categorical and Continuous Variables
- Bayesian Target Encoding
- Encodings Inspired from NLP
- Other Types of Embeddings
- week01:
- Interpretability vs Explainability
- 'Explainability' in Healthcare Applications
- Taxonomy of Explainability Methods
- Model Agnostic Explainability Methods
- Permutation Feature Importance in Time Series Data
- week02:
- Local Interpretable Model Agnostic Explanations (LIME)
- LIME in Time-Series Classification
- Shapley Additive Explanations
- Model-Specific Explanations: Visualisation Methods
- CAM in Time-Series Classification
- week03
- Gradient Weighted Class Activation Maps
- Grad-CAM in Time-Series Classification
- Integrated Gradients
- Integrated Gradients in Time Series Classification
- week04
- Attention in Deep Learning
- Taxonomy of Attention
- Attention and Explainability
- week01:
- From Reproducibility to Generalisability
- A Guide to Model Validation in Clinical Decision Support Systems
- Calibration of Deep Learning Models
- week02:
- Assessment of the Risk of Bias in EHR
- Fairness in Machine Learning for Healthcare Applications (Part 1)
- Fairness in Machine Learning for Healthcare Applications (Part 2)
- week03
- Decision Curve Analysis
- Human-Centered Clinical Decision Support Systems
- Evaluation of Explainability Models
- week04
- Privacy Concerns in CDSS
- Defences Against Inference Attacks
- Adversarial Attacks - Explainability
- Hester Huijsdens for her meticulous work in adopting, simplifying and improving related python notebooks
- Yola Jones, Shourya Verma and Hester Huijsdens for their work on explainable machine learning models
- Samuel Leighton for his insight in Clinical Decision Support Systems
- D. Ravi, C. Wong, F. Deligianni, M. Berthelot, J. Andreu-Perez, B. Lo, G. Z. Yang, Deep Learning for Health Informatics, IEEE Journal of Biomedical and Health Informatics, 21(1), 2017.
- J. Andreu-Perez, F. Deligianni, D. Ravi, G. Z. Yang, Artificial Intelligence and Robotics, UK-RAS White Papers, 2017.
- Y. Jones, F. Deligianni and J Dalton, Improving ECG Classification Interpretability Using Saliency Maps, IEEE BIBE, 2020. (Best paper award)
- M Podjaski and F. Deligianni, ‘Towards Explainable, Privacy-Preserved Human-Motion Affect Recognition’, IEEE-SSCI 2021, https://arxiv.org/abs/2105.03958.
- Leighton et al. Development and validation of multivariable prediction models of remission, recovery, and quality of life outcomes in people with first episode psychosis: a machine learning approach, The Lancet: Digital Health, 2019.
- Leighton et al. Development and validation of a non remission risk prediction model in First Episode Psychosis: An analysis of two longitudinal studies, BJPsych Open, 2021.
- Lee et al. Prediction models in first episode psychosis: a systematic review and critical appraisal, British Journal of Psychiatry, 2022.