Future Events:
The next monthly seminar will be held April 17, 2024.
Past Events:
DSI March Monthly Seminar
Speaker: Murtada K. Elbashir
Title: Identification of Hub Genes Associated with Breast Cancer Using Integrated Gene Expression Data with Protein-Protein Interaction Network
Description:
Breast cancer (BC) is the most incident cancer type among women. Early detection and prediction of BC are significant for prognosis and in determining the suitable targeted therapy. Early detection using morphological features poses a significant challenge for physicians. It is important to develop computational techniques to help determine informative genes, and hence help diagnose cancer in its early stages.
DSI February Monthly Seminar
Speaker: Justine Nasejje
Title: Time to Event Data Analysis with Machine Learning
Description:
How to leverage machine learning’s capabilities to improve the understanding, prediction, and interpretation of time-to-event outcomes, helping in risk assessment, decision-making, and prognosis in healthcare.
DSI November Monthly Seminar
Speaker: Kaleab Baye and Ramadhani Abdallah Noor
Title: Climate Change, extreme heat, and child growth in Africa
Description:
This presentation aims to highlight the various ways in which climate change can impact nutritional outcomes in Africa. It will illustrate how exposure to extreme heat during early life can increase the risk of poor child growth and nutritional outcomes. Additionally, we will discuss the implications of more frequent and prolonged exposure to extreme heat as a consequence of climate change.
DSI October Monthly Seminar
Speaker: Dr. Margaret Kruk
Title: Data for high quality health systems: what do we know, what do we need to know?
Description:
Introduction to the concepts and measures for high quality health systems and recent data on innovations.
DSI September Monthly Seminar
Speaker: Valentina Carducci, Santiago Romero-Brufau and Bowen Gu
Title: Automating Data Abstraction of Unstructured Clinical Documents for the Generation of Clinical Registries
Description:
It is estimated that around 80% of all clinical data is unstructured, free text written in natural language. This is one of the main limitations to the use of clinical data for quality improvement and research. A solution to mitigate the time, effort, and money to manually extract potentially meaningful variables that are captured in a non-structured way in free-text clinical documents has been the creation of clinical registries. These registries can be used to ask questions from the data in multiple studies, as well as drive clinical decisions, thus sharing the cost of data extraction across several research or quality-improvement initiatives. However, clinical registries are extremely expensive to create and maintain, as they often require manual extraction by data extractors (dedicated extractors, nurses, residents or consultants). The process is also quite slow, limiting how many registries can realistically be maintained, and all variables are dependent on the quality of data extraction and entry into the registry. Using AI and data engineering, we are working on automating the data abstraction process to create and maintain clinical registries more efficiently and effectively, which we think will have a huge impact in optimizing time and resources in clinical practices.
DSI April Monthly Seminar
Speaker: Dr. Isabel Madzorera
Title: Diet quality in low- and middle-income countries: Measurement, the role of food systems and implications for health outcomes
Description: Women and children on the African continent bear a significant burden of poor health and nutrition outcomes, and health disparities. The seminar will highlight the role of poor diet quality and food systems, as possible determinants of poor health outcomes for women, children and adolescents in Sub-Saharan Africa (SSA). The presentation will also discuss some of the factors influencing food systems and diet quality in SSA, including the impact of COVID-19. Finally, the seminar will highlight policy implications and strategies to improve food systems and increase the availability and access to quality diets to improve health outcomes on the African continent.
DSI March Monthly Seminar
Speaker: Dr. Mary Akinyemi
Title: Healthcare access assessment in an underserved community using a Health Marketplace Chatbot
Description: This work outlines efforts to analyze and classify healthcare experience using a marketplace chatbot for underserved community in Nigeria. The chatbot was used to access overall access to healthcare and gender bias. The primary modality of the chatbot is multiple-choice, with the bot prompting users to enter information, acknowledge continuation, or navigate from limited selections. However, users frequently attempt to take initiative through natural language expressions. We developed a set of corpora of the user initiative attempts, annotated the data with intent category labels, train and evaluate classifiers to recognize those intents on a held-out test set.
DSI February Monthly Seminar
Speaker: Laura K. Povlich
Title: Harnessing Data Science for Health Discovery and Innovation in Africa (DS-I Africa)
Description: The National Institutes of Health launched the DS-I Africa program in September 2021 with the goal to spur new health discoveries and catalyze innovation in healthcare, public health, and health research on the continent through application of data science. This talk will review the goals, structure, and progress of the program, along with potential opportunities for researchers.
DSI November Monthly Seminar
Speaker: Ashenafi Argaw Yirga
Title: Application of quantile mixed-effects model in modeling CD4 count from HIV-infected patients in KwaZulu-Natal South Africa
DSI October Monthly Seminar
Speaker: Mohanad Mohammed
Title: Statistical and Machine learning Models for Cancer Genomics Data with Applications
Description: The presentation will shortly introduce cancer as one of the health issue in the world, genomics data types, its importance and trends. In addition, applications of using statistical and machine learning models to model genomic data.
DSI September Monthly Seminar
Speaker: Dr. Deogratias Mzurikwao
Title: Building towards Explainable Artificial Intelligence (XAI)
Description: The presentation will introduce Explainable AI (XAI) as a crucial part in building responsible AI, its importance and its real life applications with reference to MUHAS emerging technologies for healthcare (mETH) Research and Development laboratory. Dr. Mzurikwao will further present about the gaps existing and potential ways to address them in order to realize the positive advantages of AI on the continent.
DSI August Monthly Seminar
Speaker: Dr. Andrew Boulle, Professor of Public Health Medicine, Western Cape Department of Health & Wellness and the University of Cape Town
Title: “Consolidated data environments for person-level health in African setting”
Description: The presentation will detail the journey towards and current functioning of the Western Cape Provincial Health Data Centre, the data science competencies which are most utilized, and the lessons learned which might be applicable to similar settings.
DSI July Monthly Seminar
Speaker: Dr. Marcello Pagano, Professor of Statistical Computing at Harvard T.H. Chan School of Public Health
Title: “What is Health Data Science?”
Description: We are launching a new program in Health Data Science. We define, together with others, health data science as a coming together of three disciplines: computing, biostatistics and health sciences. All three exist and progress well on their own but combining the three allows us the opportunity to stress what each can learn from each other to proceed synergistically and introduce an exciting new program. This talk will present a 30,000-foot view of the thematic challenges, and possible solutions, facing the launching of this progra.
DSI June Monthly Seminar
Speaker: Dr. Till Bärnighausen, Alexander von Humboldt University
Professor and Director of the Heidelberg Institute of Global Health
Description: Till’s research focuses on establishing the causal impact of global health interventions on population health, social and economic outcomes. In particular, he works on large-scale population health interventions for HIV, diabetes, hypertension, and vaccine, preventable
diseases. Till uses design research to develop interventions and randomized controlled experiments and quasi-experiments to establish intervention impacts. He has developed several new methods for applied population health research.
DSI Africa Launch
Thursday, 26th May 2022
Agenda: The proposed programme included a welcome and keynote address, introduction and overview of the grant, landscape of Data Science in our context, overview and goals of the training programme, an overview of the Health Systems Strengthening Domain, description of the Domains: Food Systems & Climate Change and Planetary Health, conversations on building Data Science Capacity in Nigeria, Ghana, Uganda, and Tanzania, a vision for Africa-wide efforts with S-S and S-N linkages, and a closing message from UKZN.