Health Data Science MSc

Using health data for research can help us to better understand the causes, prevalence, symptoms, and treatment of disease, as well as understanding how to improve health care systems. Health data science can improve our knowledge of health and care, and is an emerging discipline, arising at the intersection of statistics, computer science, and health. The health data science team can generate data-driven solutions through comprehension of complex real-world health problems, employing critical thinking and analytics to derive knowledge from data. This module will provide an understanding of the potential benefits and challenges in the application of data science to healthcare. Students will benefit from research-connected teaching with academic staff, and have opportunities for active learning including communication skills. The module will be taught via formal lectures, seminars from guest speakers, and practical communication sessions. Learning will be assessed via a coursework and a practical assessment.

Statistics is the science concerned with the collection, analysis, interpretation and presentation of data to extract knowledge. Understanding key statistical concepts is fundamental to the health data scientist. In this module students will be introduced to the concepts of variability and sampling and the different paradigms for statistical inference. The essential skills of reading data, structuring data, conducting statistical analyses and the importance of data visualisation will be covered to gain an in-depth knowledge and understanding of statistical methods used in the analysis and presentation of health data. Students will benefit from research-connected teaching with academic staff, and have opportunities for active learning. The module will be taught via formal lectures, seminars from guest speakers, and computer practical sessions. Learning will be assessed via a poster presentation and a practical assessment.

The volume of data generated by modern healthcare and public health systems is immense. Unfortunately, this data is often disconnected, carries biases, and is encoded with diverse disease ontology frameworks. However, the opportunity to transform this reality is within our grasp. We can forge a more interconnected and fair health system through collaboration and collective effort. This module offers research-focused teaching, empowering students to effect change and emerge as leaders in the field. The curriculum comprises formal lectures, seminars from guest speakers, and hands-on computer lab sessions, ensuring students are well-prepared for the challenges ahead. Through a practical evaluation and a pre-recorded oral poster presentation, students will be equipped to make a tangible impact on real-world healthcare and public health challenges. Learning will be assessed via a practical assessment, and a poster and pre-recorded oral presentations.

Independent research is the defining feature of a postgraduate student. In this module, students will conduct a mix of applied and methodological research study under the supervision of one member of staff from Biostatistics, Computer Science or Public Health, with the potential addition of a domain expert or Health Care Professional if it is required by the subject of the dissertation. Students will identify the research question and use appropriate methodologies to answer a specific gap in existing knowledge in health data science. There will be a minimum of 12 hours of supervisory meetings to assist each student in achieving this. The written assessment is a 10,000 word dissertation.

This module focuses on how databases are used in modern information systems. They are at the heart of almost all systems, such as supermarket checkouts, online banking, home rentals, and much more. One of the most successful data definition and manipulation languages is SQL, which will be covered in detail. The module will also introduce some of the fundamental concepts in computer science, as well as the mathematical underpinnings of relational databases and the techniques used to support concurrency and reliability in large information systems.

This module is intended to explore qualitative research methods in a holistic manner; moving from research philosophy, through design to individual research methods and analysis. The module covers a range of qualitative research methods through a mixture of lectures and workshops. In undertaking this module students will consider how research design and individual research methods need careful selection to suit the specific research problems or questions under investigation.

Digital technology offers great potential for improving the design, conduct and analysis of studies evaluating health care interventions. Recent evidence shows the utility of long-term follow-up of clinical trial patients through the electronic health record. Information collected directly from trial participants, through wearables, apps, and online patient-reported outcome measurement, can supplement routinely collected clinical data. Searching electronic health records for eligible patients that could benefit from a particular trial may improve the assessment of feasibility of trial recruitment and address known challenges. The aim of this module is to provide an awareness of how today’s technology could improve the efficiency of evaluations of health care interventions, and where further improvements are needed. Students will benefit from research-connected teaching with academic staff, and have opportunities for active learning. The module will be taught via formal lectures, seminars from guest speakers, and discussion groups. Learning will be assessed via a coursework and a practical assessment.

This module will provide students with the informatics knowledge and skills to turn health data into information that can be actioned in health systems, producing improved care, better data and better policies – a so called Learning Health System. It will bring together data science and informatics engineering approaches to building technical and human systems that underpin both research and business intelligence for health systems. Students will learn about the core elements of Learning Health Systems and the principles of developing and evaluating an informatics intervention. Students will benefit from research-connected teaching with academic staff, and have opportunities for active learning. The module will be taught via formal lectures, seminars from guest speakers, and computer practical sessions. Learning will be assessed via an oral presentation in which students will communicate actionable findings on a close-to-reality synthetic dataset, and a written analysis report on six-week collaborative Data Lab exercise.

Independent research is the defining feature of a postgraduate student. In this module, students will conduct a mix of applied and methodological research study under the supervision of one member of staff from Biostatistics, Computer Science or Public Health, with the potential addition of a domain expert or Health Care Professional if it is required by the subject of the dissertation. Students will identify the research question and use appropriate methodologies to answer a specific gap in existing knowledge in health data science. There will be a minimum of 12 hours of supervisory meetings to assist each student in achieving this. The written assessment is a 10,000 word dissertation.

Healthcare data consists of a range of different data types, such as blood pressure over time, time until death, or family history of a disease. The goal is usually to make predictions for the future to facilitate more informed treatment choices, improve patient care, and raise patient quality of life. However, simple approaches remain widespread amongst researchers leading to inefficient analysis and presentation of results, and therefore potential for important predictive relationships to be missed. The aim of this module is to provide an awareness of the potential issues occurring in data from clinical and observational settings, and teach a variety of techniques to conduct statistical analyses for the estimation and clinical prediction using real world data. Students will benefit from research-connected teaching with academic staff, and have opportunities for active learning. The module will be taught via formal lectures, seminars from guest speakers, and computer practical sessions. Learning will be assessed via a Statistical Analysis Plan, oral presentation, and written report.

Real world health data often has complex structures, which can impact on individual health outcomes. Outcomes may depend on existing administrative or geographic structures, or the risk of developing disease may be based on complicated combinations of underlying factors. With growing access to high-dimensional datasets, both in health research, and across the data-science spectrum, suitable statistical methods are essential to harness the information within these datasets. This module will provide an understanding of how appropriate statistical methods can be selected, and teach the skills necessary to conduct analyses on real world data. In this module, students will learn a variety of statistical modelling and learning algorithms, and be introduced to state-of-the-art machine learning methods for data mining and classification. The tools developed during this module will provide students with the understanding and skills needed to perform complex analysis as part of a data science team in their future careers. Students will benefit from research-connected teaching with academic staff, and develop hands-on experience of data analysis in data-labs.

In today’s healthcare landscape, understanding which genomic factors influence disease risk and treatment response is paramount. Our module delves into the pivotal role of genomic data in deciphering disease aetiology and tailoring treatment plans to individuals, aligning with the ambition of healthcare providers globally to integrate genomics into patient care. Through a blend of theoretical teaching and hands-on exercises, you’ll gain proficiency in specialized statistical methods and programming necessary to navigate complex genomic datasets. From genotype quality control to polygenic risk scores, our comprehensive curriculum covers key analysis techniques essential for genomic research, including both traditional statistical techniques and machine learning methods. It assumes no prior knowledge of genomic data, with an introduction to the terminology and structure of the data, as well as to linux programming during the first two weeks. With a focus on practical application, and supplemented by guest lectures from leading experts, our module prepares you to be able to confidently analyse and interpret the huge and complex datasets that are typical within statistical genetics and pharmacogenomics research. Learning will be assessed via two practical assessments.

​Biologically inspired optimisation and introduction to neural networks for artificial intelligence.

The module covers a range of topics and techniques for analyzing data. Students will learn about different types of data mining problems, including classification, clustering, association pattern mining, and social network analysis, as well as algorithms to solve them.

Students will program selected data mining algorithms from scratch using Python. This hands-on approach will allow them to gain a deeper understanding of how the algorithms work and how they can be applied to real-world datasets. They will experiment with different datasets to see how the algorithms perform and learn how to interpret the results.

This module teaches you about bio-inspired algorithms for optimisation and machine learning. The algorithms are based on reinforcement learning, DNA computing, brain or neural network models, immune systems, the evolutionary version of game theory, and social insect swarm behaviour such as ant colonies and bee colonies. These techniques are extremely useful for searching very large solution spaces (optimisation) and they can be used to design agents or robots that have to interact and operate in dynamic unknown environments (e.g. a Mars rover, a swarm of robots or network of satellites). The idea of learning optimal behaviour, rather than designing, algorithms and controllers is especially appealing in AI.  ​

Independent research is the defining feature of a postgraduate student. In this module, students will conduct a mix of applied and methodological research study under the supervision of one member of staff from Biostatistics, Computer Science or Public Health, with the potential addition of a domain expert or Health Care Professional if it is required by the subject of the dissertation. Students will identify the research question and use appropriate methodologies to answer a specific gap in existing knowledge in health data science. There will be a minimum of 12 hours of supervisory meetings to assist each student in achieving this. The written assessment is a 10,000 word dissertation.