This page gathers the stroke modelling work for the Stroke Audit Machine Learning (SAMueL) project.
An overview of the project and a list of related publications are available on the NIHR website.
[Click here] What problem are we addressing?
[Click here] What did we do?
[Click here] What did we find?
[Click here] What did doctors think?
Our key results are published on these online books:
SAMueL-1 project (complete)
These books cover the whole process of finding our results. They contain both plain-English summaries and more technical walkthroughs.
The contents include background information on stroke and treatment, introductions to the methods and concepts used, and applications of the methods to our data.
Some of the SAMueL-2 work is also shared between other projects, and some code will be packaged up and stored elsewhere. The following organisations will store that content when it is ready:
 stroke-modelling |
 |
Python packages for easily using our models. |
 stroke-optimist |
|
Stroke OPTIMIST Project: OPTimising IMplementation of Ischaemic Stroke Thrombectomy. |
 stroke-digital-twin |
|
Digital twins of the stroke pathway. |
As of April 2024, the public repositories here are the following:
List of tags: 💻 Modelling 🧮 Data prep 📜 Paper 🖼️ Slides 🍃 Overleaf 🎮 Streamlit app 📎 Admin 🕮 Online book 🧪 Test ⚖️ SHAP 🔮 Machine learning 📋 Outcomes ⏱️ Pathway 🗺️ Geography
[Click here] Code repository list
| Repository | Description | Tags |
|---|---|---|
| samuel_2_production | Core code for SAMueL-2 | 💻 Modelling |
| ssnap_production_code | Code for running of SAMueL analysis by SSNAP | 💻 Modelling |
| thrombolysis_organisational_factors | How do organisational factors affect thrombolysis? | 💻 Modelling |
| stroke_outcome_ml | Predicting the discharge disability of stroke patients | 💻 Modelling⚖️ SHAP🔮 Machine learning📋 Outcomes |
| geography_data | Data prep for geography - LSOA locations, regions, shape files etc. | 🧮 Data prep |
| samuel_example | Replication of Stroke Audit Machine Learning with artificial patient data | 💻 Modelling⚖️ SHAP🔮 Machine learning |
| stroke_outcome_xgb_shap_TO_ARCHIVE | XGB model, with SHAP, for stroke outcome | 💻 Modelling⚖️ SHAP🔮 Machine learning📋 Outcomes |
| skeleton-pathway-model | Skeleton SimPy stroke pathway model from onset to thrombolysis and thrombectomy | 💻 Modelling⏱️ Pathway |
| stroke_outcome | Outcome modelling | 💻 Modelling📋 Outcomes |
| samuel_causal | Causal analysis and diagrams for the SAMueL project | 💻 Modelling |
| synthetic_data | Create synthetic data from SAMueL data | 💻 Modelling🧮 Data prep |
| model_comparison | A comparison of different model types using SAMueL-1 data | 💻 Modelling🔮 Machine learning |
| stroke_unit_demographics | Collating demographic data for emergency stroke unit catchment areas | 🧮 Data prep |
| samuel_2_data_prep | SAMUeL_2 data preparation | 🧮 Data prep |
| streamlit_combo_stroke | Combined the existing stroke streamlit apps into one multipage app | 🎮 Streamlit app⚖️ SHAP🔮 Machine learning📋 Outcomes⏱️ Pathway |
| streamlit_pathway_improvement | Streamlit app for pathway improvement data | 🎮 Streamlit app📋 Outcomes⏱️ Pathway |
| streamlit_stroke_treatment_ml | Streamlit app for machine learning model to predict treatment given to emergency stroke patients | 🎮 Streamlit app⚖️ SHAP🔮 Machine learning |
| streamlit_descriptive_stats | Streamlit app for descriptive statistics for each stroke team in the SAMuEL project | 🎮 Streamlit app |
| stroke_outcome_app | Streamlit app for stroke outcome modelling | 🎮 Streamlit app📋 Outcomes |
| streamlit_map_lsoa_outcomes | Test app for maps in streamlit | 🎮 Streamlit app🧪 Test📋 Outcomes🗺️ Geography |
| causal_inference_basics | Basics of causal inference | 🧪 Test |
| smote-variation | Variation of SMOTE | 🧮 Data prep🧪 Test |
| import_from_relative_path | Demo to show how to import a module from a package in a different directory | 🧪 Test |
[Click here] Papers and presentations repository list
| Repository | Description | Tags |
|---|---|---|
| samuel-2-reference | A repository of general reference documents for the SAMueL-2 project | 📎 Admin |
| samuel-1 | (blank) | 🕮 Online book |
| samuel-2 | Jupyter book for SAMueL-2 project | 🕮 Online book |
| .github | For this organisation's README etc. | 📎 Admin |
| overleaf_stroke_outcome_1 | Open paper on stroke outcome modelling | 🍃 Overleaf📜 Paper📋 Outcomes |
| stroke_treatment_review | overleaf_stroke_treatment_review | 🍃 Overleaf 📜 Paper |
| overleaf_samuel_shap_presentation | SHAP presentation | 🍃 Overleaf🖼️ Slides⚖️ SHAP🔮 Machine learning |
| overleaf_shap_paper_2 | SHAP paper focusing on interactions | 🍃 Overleaf📜 Paper⚖️ SHAP🔮 Machine learning |
| overleaf_shap_paper_1_for_esj | Overleaf_SHAP_paper_1_for_ESJ | 🍃 Overleaf 📜 Paper⚖️ SHAP🔮 Machine learning |
| overleaf_shap_paper_1_short | Overleaf SAMueL SHAP Paper 2 | 🍃 Overleaf📜 Paper⚖️ SHAP🔮 Machine learning |
| overleaf_samuel_1_contentious_patients | Paper | 🍃 Overleaf📜 Paper⚖️ SHAP🔮 Machine learning |
| overleaf_shap_pci_jan_2023 | Patient and carers meeting Jan 2023 | 🍃 Overleaf🖼️ Slides⚖️ SHAP🔮 Machine learning |
| overleaf_shap_paper_1_long | Shap paper 1 - long - preprint | 🍃 Overleaf📜 Paper⚖️ SHAP🔮 Machine learning |
| overleaf_stakeholder_cambridge_icb_dec_2022 | Presentation to the Cambridge and Peterborough Integrated Care Board (Health Inequalities) | 🍃 Overleaf🖼️ Slides🔮 Machine learning📋 Outcomes🗺️ Geography |
| overleaf_advisory_group_nov_2022 | SAMueL Advisory Group November 2022 | 🍃 Overleaf🖼️ Slides⚖️ SHAP🔮 Machine learning📋 Outcomes |
| overleaf_samuel_overview | Overleaf beamer slides for an overview of SAMueL, originally made for an HSMA talk in November 2022. | 🍃 Overleaf🖼️ Slides🔮 Machine learning📋 Outcomes🎮 Streamlit app |
| overleaf_coproduction_workshop_1 | Coproduction workshop slides | 🍃 Overleaf🖼️ Slides⚖️ SHAP🔮 Machine learning |
| overleaf_samuel_pci_oct_2022 | pci slides | 🍃 Overleaf🖼️ Slides |
| samuel_shap_paper_2 | Continuing exploratory work with Shap using SAMueL-1 data | 📜 Paper⚖️ SHAP🔮 Machine learning |
| samuel_shap_paper_1 | Exploratory work with Shap using SAMueL-1 data | 📜 Paper⚖️ SHAP🔮 Machine learning |