Healthcare’s growing use of data is changing the expertise required to advance research and improve care. For example, the analysis of a tumour may involve pathology images, molecular measurements and clinical records. Understanding a patient’s risk may require information collected across several healthcare encounters. Turning these different sources into useful evidence calls for closer interaction between clinical knowledge, biological science and computational methods.
Medical AI, bioinformatics and health data science are central to this evolution. While each has its own methods and areas of application, they increasingly intersect around shared questions about disease, treatment and healthcare delivery. For professionals, this creates opportunities to extend established expertise into new areas of research, evaluation and implementation.
The UAE provides a relevant setting in which to examine these changes. The Emirati Genome Programme and associated precision-medicine initiatives connect genomic research with questions in oncology, pharmacogenomics and rare diseases. Abu Dhabi’s Malaffi uses AI-based predictive models to assess patient risks using information such as diagnoses, laboratory results and medications. These initiatives illustrate the range of expertise involved in connecting biological and clinical information with healthcare decisions.
Different disciplines, connected questions
Medical AI focuses on computational methods that address medical questions. Applications include interpreting scans and tissue images, assessing risk and supporting clinical decisions. Developing these systems requires expertise in programming, statistics and machine learning, alongside an understanding of how the clinical problem has been defined.
For example, a model designed to identify disease in an image must be evaluated in relation to the consequences of missed cases and incorrect findings. Its usefulness also depends on where it enters the clinical workflow and what action follows its output. Clinicians, researchers, engineers and implementation specialists therefore contribute different forms of knowledge to the same application.
Bioinformatics centres on the analysis of biological data, including genomic sequences, gene expression and protein measurements. It helps researchers investigate disease mechanisms, identify potential biomarkers and explore differences in treatment response. Interpretation depends on understanding both the analytical method and the biological processes that produced the data.
A laboratory scientist may recognise how sample preparation affects a result, while a computational specialist may identify limitations in the analysis. Clinical expertise helps establish whether a finding addresses a meaningful medical question. Progress depends on these perspectives informing one another.
Health data science examines questions using clinical records, population data and information about healthcare services. It can support the study of patient outcomes, disease patterns, resource use and the effectiveness of interventions. Here, understanding how records are generated is particularly important. Missing information, inconsistent coding and differences between patient groups can influence what an analysis appears to show.
Expertise across the research process
These disciplines often meet within a single project. A cancer study might combine tissue images with molecular findings and treatment histories to investigate why outcomes differ between patients. Before analysis begins, the team must establish whether the information is comparable, whether outcomes have been defined consistently and whether the available data can answer the research question.
The work continues after a model or analysis produces a result. Findings need to be examined for alternative explanations, tested in relevant populations and interpreted in relation to existing knowledge. A technically strong result may still have limited clinical significance.
This creates a broader range of professional contributions than model development alone. Data preparation, analytical pipeline development, clinical evaluation, biological interpretation and workflow integration all require specialist judgement. Professionals who can communicate across these activities help teams maintain a coherent connection between the original question and the conclusion.
Building on existing professional strengths
For clinicians, participation in medical AI may begin with defining a useful application or evaluating a proposed tool. Their knowledge of patient care helps determine which outcomes matter and whether an application fits the circumstances in which decisions are made. Familiarity with sensitivity, specificity, uncertainty and bias supports more critical assessment of performance claims.
Biomedical and laboratory professionals can develop towards bioinformatics through training in statistics, programming and the supervised analysis of relevant datasets. Their understanding of experimental methods provides an important foundation for recognising when a computational finding is biologically plausible.
Computing professionals bring expertise in software, algorithms and data systems. Developing an understanding of clinical workflows, research design and healthcare-data limitations enables them to apply those skills more effectively. Professionals in public health and healthcare operations can contribute through population-health analysis, service evaluation and the study of outcomes.
Across these pathways, disciplinary depth remains important. The aim is to develop enough understanding of adjacent fields to collaborate effectively, recognise limitations and ask better questions.
Learning through practical contribution
Professional development is most useful when learning is connected to a defined responsibility. Improving the quality of a dataset, reproducing a published analysis or contributing to an evaluation can reveal challenges that are less apparent in classroom exercises. Working with colleagues from other disciplines also exposes professionals to different assumptions about what constitutes reliable evidence.
Supervised projects, research collaborations and placements can provide this experience. Their educational value depends on clear objectives, access to appropriate expertise and constructive feedback. A carefully documented project should explain the question, methods, findings and limitations, demonstrating how the professional reached and assessed a conclusion.
Short courses can address specific knowledge gaps, while formal postgraduate study may be appropriate for sustained specialist development or original research. The choice should follow the intended contribution. Someone seeking to analyse genomic data requires a different preparation from someone responsible for evaluating an AI application or leading change across a healthcare service.
Developing the conditions for meaningful work
Employers also shape how these careers develop. Access to suitable data, clinical or scientific collaborators, mentorship and time for evaluation can determine whether professionals are able to apply their training productively. Hospitals, laboratories, technology companies and universities have complementary roles in creating these conditions.
As responsibilities grow, the ability to explain uncertainty and recognise the boundaries of one’s expertise becomes increasingly important. Responsible data handling, transparent methods and clear communication are part of professional competence across all three fields.
The evolution of healthcare careers will be shaped by how effectively organisations connect these capabilities. Medical AI, bioinformatics and health data science offer different ways to contribute but share a common requirement: the ability to turn complex information into evidence that can withstand scrutiny. Developing that capability will be essential to translating scientific and technological advances into better healthcare.
Prof. Kashif Rajpoot, Professor of Medical AI, University of Birmingham Dubai
