We are thrilled to announce the release of our latest software packages, specifically designed to enhance your use of machine learning technology in a wide range of healthcare topics (e.g. clinical trials, drug development, prognosis, personalised medicine). Our ground-breaking tools offer novel solutions to common and specific healthcare challenges alike and enable you to achieve new levels of efficiency, accuracy, and transparency in both research and everyday clinical work.
Whether you are a data scientist, drug developer, or practicing clinician, our software packages are designed to meet your needs and help you achieve your goals.
These tools represent part of our fundamental aims as a centre – to establish and understand challenges in healthcare in co-operation with our industry partners, and to address these issues with innovative technologies.
For 15 June, we will organise a comprehensive AI Clinic as part of the van der Schaar Lab’s Revolutionizing Healthcare engagement sessions. This event is open to all practising clinicians. You can find out more about Revolutionizing Healthcare here.
Under our new Software tab, you can find:
AutoPrognosis – an automated predictive modelling pipeline designed for clinical prognosis, leveraging state-of-the-art advances in automated machine learning to optimise ML pipelines, incorporate model explainability tools, and enable deployment of clinical demonstrators without requiring significant technical expertise.
Possible use cases: Streamlining the development of diagnostic and prognostic models and easily customising them to a given task such as cancer prognosis or clinical trials
HyperImpute – a comprehensive library for handling missing data in your ML pipelines, simplifying the selection process of a data imputation algorithm and offering a range of novel algorithms compatible with sklearn.
Possible use cases: Enabling faster and more reliable model training and using a wide range of imputation algorithms for gene expression analysis, drug discovery, personalised medicine, image analysis etc
Interpretability Suite – a comprehensive collection of Machine Learning interpretability methods, offering users a reference to select the best-suited method for their needs, with a focus on providing insights into ML model predictions through organisation, a common Python interface, and a user interface.
Possible use cases: Enhancing transparency and trust in ML models and accelerating the implementation of interpretability methods for a more efficient and faster decision-making process in treatment decisions, diagnostics etc
Synthcity – an open-source synthetic data generation library that outperforms rivals (YData, Gretel, SDV, etc.) in terms of compatible use cases and data modalities, offering solutions for privacy, data scarcity, and fairness across various data types.
Possible use cases: Addressing data privacy concerns, combat data scarcity, and facilitate rapid experimentation, prototyping, and benchmarking in clinical trials and drug development
TemporAI – a Machine Learning-centric time-series library tailored for medicine, focusing on time-series prediction, time-to-event analysis, and counterfactual inference for individualised treatment effects.
Possible use cases: Leveraging time-series data to improve clinical trial designs, enhance drug discovery, and enable personalised medicine