AI AND HEALTH - 2022/3
Module code: EEEM069
The module provides an application-focused tour of machine learning for real-world healthcare research and application from understanding various healthcare components, ethical concerns to pre-processing and analysing healthcare data for classification, survival and risk analysis, and early prediction tasks. It requires the knowledge of basic machine learning, linear algebra, and familiarity with Python programming.
Electrical and Electronic Engineering
KOUCHAKI Samaneh (Elec Elec En)
Number of Credits: 15
ECTS Credits: 7.5
Framework: FHEQ Level 7
Module cap (Maximum number of students): N/A
Overall student workload
Independent Learning Hours: 108
Lecture Hours: 33
Laboratory Hours: 9
Prerequisites / Co-requisites
Indicative content includes:
The module first introduces healthcare components and various sources of healthcare data. It then discusses the ethical concerns and various sources of bias in analysing healthcare data. These concepts then will be considered as the basis to pre-process, analyse and evaluate real-world healthcare data using various machine learning techniques. The learned concepts will be reinforced through lab sessions in Python.
33 hours of lectures:
¿ Introduction to healthcare systems and their components, layers of care, knowledge graphs and coding systems (e.g., ICD-10), electronic health and medical records, and quality measures [1-2].
¿ AI applications in delivery of health care services and ethical issues, including introduction to various healthcare applications, various sources of bias, and implications, ethical frameworks, limitations of AI on healthcare data, and second use of data [3-5].
¿ Formulating important clinical questions, different types and sources of clinical data such as clinical texts, omics data, medical imaging, signals and their values, application, and major issues [6-8].
¿ Clinical data pre-processing, including temporal information and data aggregation, medical records pre-processing (standardisation, imputation, and feature selection/extraction), and phenotyping [9-12].
¿ Clinical machine learning for healthcare, concepts, definitions, and design choices , linear / logistic regression, odds, and risk [13-14], survival analysis, hazard, and survival rate , traditional machine learning and deep learning, supervised, unsupervised, and weak/self-supervision [16-23], correlation vs causation , interpretable learning [24-25], dealing with data imbalance and missing values [26-30], and regularisation .
¿ important metrics for clinical practice, data quality vs quantity, and feasibility, impact, utility, and clinical evaluation [30-33].
9 sessions of labs:
¿ Loading various types of healthcare data and understanding their differences
¿ Dealing with missing data and data imputation
¿ Dealing with imbalance data and data imputation
¿ Analysing and evaluating machine learning models for various types of healthcare data [classification, early warning system, and survival analysis]
¿ Interpretability analysis
|Assessment type||Unit of assessment||Weighting|
|Practical based assessment||Lab Report||18|
|Examination Online||Examination ONLINE (OPEN BOOK) EXAM WITHIN 4HR WINDOW||60|
The assessment strategy is designed to provide students with the opportunity to assess all taught materials through use of broad range of questions covering problem solving questions that require recommendation of appropriate algorithms and solutions. Examination will cover all taught materials following the lecture notes. The practical assignment focuses on implementation and evaluation of a machine learning system for a real-world health care data with the focus on the selection of appropriate machine learning techniques and their implementation and evaluation.
Thus, the summative assessment for this module consists of:
¿ Written examination (60% weighting).
¿ Coursework assignment in Python (22% weighting). Set week 3, due week 9.
¿ Lab-based assignments in Python (18% weighting).
Formative assessment and Feedback
For the module, students will receive formative assessment/feedback in the following ways.
¿ During lectures by question and answer sessions
¿ By means of lab problem sheets
¿ During supervised lab sessions
¿ Via feedback comments on assessed coursework
- This module aims to introduce:
¿ healthcare systems and ethical concerns,
¿ various sources of healthcare data (e.g. electronic records, signals and clinical texts) and challenges associated with data analysis (feature selection/extraction, imputation, augmentation and standarisation),
¿ machine learning methods to analyse healthcare data for early prediction, risk analysis, diagnosis, and survival analysis,
¿ techniques to validate the performance of models in clinical practice and the importance if various performance measures,
¿ interpretability frameworks (e.g., SHAP, LIME, deep learning based approaches) and their use in healthcare.
|001||Identify ethical concerns and various sources of bias, the difference between various data sources, phenotypes, and clinical questions, and relate appropriate machine learning solutions to solve them||T|
|002||Implement various pre-processing techniques, appropriate machine learning solutions to different healthcare problems and interpretability frameworks to analyse healthcare data and identify important clinical features||CP|
|003||Recognise and use different aspects of model validation, evaluate the performance of developed models, and draw conclusion on their applicability and efficiency||T|
C - Cognitive/analytical
K - Subject knowledge
T - Transferable skills
P - Professional/Practical skills
Methods of Teaching / Learning
The learning and teaching strategy is designed to:
to deliver background and theory in lectures and use the lab sheets for practical application of the learnt theory. The latter also provides an opportunity for formative feedback. The coursework exposes students to the full development cycle of a clinically applicable system ¿ data understanding and pre-processing, implementation, evaluation, interpretability analysis and reporting of the conclusions and challenges.
The learning and teaching methods include:
¿ Lectures ¿ 3 hours per week x 11 weeks
¿ Computer labs ¿ 1 hour per week x 9 weeks (weeks 2-10).
Indicated Lecture Hours (which may also include seminars, tutorials, workshops and other contact time) are approximate and may include in-class tests where one or more of these are an assessment on the module. In-class tests are scheduled/organised separately to taught content and will be published on to student personal timetables, where they apply to taken modules, as soon as they are finalised by central administration. This will usually be after the initial publication of the teaching timetable for the relevant semester.
Upon accessing the reading list, please search for the module using the module code: EEEM069
Required purchase Essential reading  Liu, Y., Chen, P. H. C., Krause, J., & Peng, L. (2019). How to read articles that use machine learning: users¿ guides to the medical literature. Jama, 322(18), 1806-1816. https://jamanetwork.com/journals/jama/fullarticle/2754798  Kelly, C.J., Karthikesalingam, A., Suleyman, M. et al. Key challenges for delivering clinical impact with artificial intelligence. BMC Med 17, 195 (2019). https://bmcmedicine.biomedcentral.com/articles/10.1186/s12916-019-1426-2  Johnson, Alistair EW, et al. "MIMIC-III, a freely accessible critical care database." Scientific data 3.1 (2016): 1-9. Recommended reading  Topol, Eric. Deep medicine: how artificial intelligence can make healthcare human again. Hachette UK, 2019.  Bohr, Adam, and Kaveh Memarzadeh, eds. Artificial intelligence in healthcare. Academic Press, 2020.  Gerke, Sara, Timo Minssen, and Glenn Cohen. "Ethical and legal challenges of artificial intelligence-driven healthcare." Artificial intelligence in healthcare. Academic Press, 2020. 295-336.  Chen, Yen-Wei, and Lakhmi C. Jain. Deep Learning in Healthcare. Cham: Springer, 2020.  Goldstein, B.A., Navar, A.M., Pencina, M.J. and Ioannidis, J., 2017. Opportunities and challenges in developing risk prediction models with electronic health records data: a systematic review. Journal of the American Medical Informatics Association, 24(1), pp.198-208.  Hernandez-Boussard, T., Tamang, S., Blayney, D., Brooks, J. and Shah, N., 2016. New paradigms for patient-centered outcomes research in electronic medical records: an example of detecting urinary incontinence following prostatectomy. eGEMs, 4(3).  Matheny, M., Israni, S. T., Ahmed, M., & Whicher, D. (2020). Artificial intelligence in health care: The hope, the hype, the promise, the peril. Natl Acad Med, 94-97. https://nam.edu/artificial-intelligence-special-publication/  Wang, F., Kaushal, R., & Khullar, D. (2020). Should health care demand interpretable artificial intelligence or accept ¿black box¿ medicine? https://www.acpjournals.org/doi/10.7326/M19-2548 Background reading  Bishop, Christopher M., and Nasser M. Nasrabadi. Pattern recognition and machine learning. Vol. 4. No. 4. New York: springer, 2006.  Goodfellow, Ian, Yoshua Bengio, and Aaron Courville. Deep learning. MIT press, 2016.
Please note that the information detailed within this record is accurate at the time of publishing and may be subject to change. This record contains information for the most up to date version of the programme / module for the 2022/3 academic year.