Tag: Biomedical AI

Turning Evidence Into Action: The AEQUITAS Database

AI tools used in healthcare can carry hidden biases, sometimes performing less accurately for women, ethnic minorities, or other underrepresented groups. But knowing this risk exists in general is one thing. Being able to check whether a specific tool, in a specific clinical area, has documented bias issues is another, and that’s exactly the gap the AEQUITAS Database was built to fill.

The AEQUITAS Database is a structured, digital repository that systematically collects, organises, and provides access to evidence on gender and racial bias in biomedical AI. Rather than leaving healthcare professionals, researchers, and policymakers to piece together scattered studies and reports, it brings this dispersed evidence together into one coherent, accessible knowledge base.

The database is publicly accessible through the AEQUITAS project website, and it was built with a user-oriented approach, meaning both technical and non-technical users can navigate it effectively.

Built for multiple audiences

The Database was designed to serve several groups at once, each with different needs. For healthcare professionals, it’s a resource for identifying potential biases in AI-based diagnostic and decision-support tools, improving patient-centred care, and reducing the risk of misdiagnosis. For civil society organisations, it supports awareness-raising and advocacy for fair, inclusive healthcare systems. And for researchers and policymakers, it feeds into further scientific investigation, the development of clinical guidelines, and regulatory oversight at European level.

A practical tool

The Database is designed to be used, not just consulted occasionally. Before adopting a new AI tool, healthcare professionals can search for entries related to its disease area to check whether bias has been documented in similar tools. When an AI system produces unexpected outputs, the Database can help determine whether it’s a known pattern rather than an isolated anomaly. And when caring for patients from underrepresented groups, it can help clinicians understand whether the tools involved in their care carry documented disparities relevant to that patient’s profile.

The Database is also a living resource that will grow as new evidence becomes available. The more healthcare professionals, researchers, and organisations engage with it, the more valuable it becomes for everyone.
From Principles to Practice: How AEQUITAS Turns Policy into Action

AI systems used in healthcare are governed by strict ethical principles and legal requirements, from the EU AI Act to the EU Charter of Fundamental Rights. But knowing the rules isn’t the same as knowing how to apply them. Hospitals and civil society organisations (CSOs) need concrete, operational processes they can actually put into practice, day to day, to make sure these tools are used safely and fairly.

This is exactly what the AEQUITAS AI Regulatory Model was designed to do. It gives hospitals and CSOs a practical governance framework covering the full lifecycle of AI systems, from procurement and validation to deployment, monitoring, and accountability.

The model draws inspiration from approaches that have proven effective in other high-risk sectors. It echoes the logic of pharmacovigilance, the systematic tracking of adverse drug reactions, in what AEQUITAS calls an “AI-vigilance” system: a way to report and analyse incidents involving AI-supported clinical decisions, with a particular focus on equity outcomes. It also mirrors the patient safety culture familiar from aviation-inspired approaches in hospitals, built around identifying hazards, reporting incidents, and continuously improving.

At its core, the model is grounded in the EU Charter of Fundamental Rights (2009), particularly the principles of human dignity, equality and non-discrimination, and the right to healthcare, translating these rights into operational safeguards that follow an AI system throughout its use.

The five checkpoints

Rather than treating AI governance as a single box to tick, the model structures oversight around five sequential stages, each one a checkpoint designed to catch potential equity risks before they reach patients. Throughout this process, hospitals act as the main deployers, leading decisions and operational control, while CSOs play an independent oversight role, reviewing documentation, flagging equity-related risks, and conducting audits, without controlling day-to-day clinical decisions themselves.

These stages follow the natural lifecycle of any AI system in a hospital. It starts with scope and classification, determining an AI tool’s regulatory status before it’s even procured or developed. This sets the direction for everything that follows. Next comes procurement, often the moment hospitals have the greatest leverage, as it’s where suppliers can be required to commit to addressing bias risks throughout a system’s lifecycle. At this stage, CSOs are able to review procurement documentation and flag equity concerns. The third stage, validation, means independently evaluating how a system performs, not just overall, but across different patient groups, before it’s ever used with real patients.

Once a system passes validation, the deployment stage ensures the right safeguards are in place: trained oversight staff, mechanisms for clinicians to override AI recommendations, and clear communication with patients about when these tools are being used in their care. And because AI systems aren’t static, the final stage, ongoing monitoring, auditing, and accountability, keeps watching for emerging disparities long after a system goes live, with CSOs conducting independent audits and helping escalate persistent issues.

Building a shared network across Europe

Beyond the model itself, AEQUITAS is currently establishing a European network of CSOs and public hospitals. This shared governance layer means that insights, incidents, and mitigation strategies identified in one country contribute to a growing, Europe-wide knowledge base, one that benefits patients well beyond the institution where the issue was first spotted.

Together, these tools turn abstract principles into something hospitals and CSOs can actually act on, closing the gap between good intentions and equitable, everyday care.

References:

European Union. (2009). Charter of Fundamental Rights of the European Union. Official Journal of the European Union, C 303/1. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:12012P/TXT
Beyond Ethics: How Europe Is Regulating AI in Healthcare

Knowing that AI can be biased is one thing. Knowing what to do about it, legally and ethically, is another.

Medicine’s ethical foundations, extended to AI

Medicine has long been associated with high ethical standards, from the Hippocratic Oath to the Declarations of Geneva and Helsinki. Unfortunately, the history of medicine includes deeply troubling chapters, from eugenics, forced sterilisation of indigenous and marginalised populations, to non-consensual experimentation on people with disabilities. These historical failures are directly relevant today because the structural injustices behind them shaped decades of medical data collection, and their legacy is exactly what today’s AI systems risk inheriting when trained on that same data.

Biomedical ethics is built around four core ethical principles: autonomy (respecting a person’s right to make their own decisions), non-maleficence (avoiding harm), beneficence (actively promoting wellbeing), and justice (distributing benefits and risks fairly, especially given known disparities based on race, gender, and social status) (Beauchamp & Childress, 2019).

As AI entered clinical practice, ethicists added a fifth principle specifically for these technologies: explicability, the idea that the reasoning behind an AI system’s decisions must be understandable and open to scrutiny, not simply accepted at face value (Floridi et al., 2018).

The EU’s legal response

The European Union has built a multi-layered legal framework specifically to address these risks. At its centre is the AI Act, which classifies many healthcare applications, including diagnostic tools and clinical decision-support systems, as “high-risk.” This means they’re subject to strict requirements: providers must actively identify and mitigate risks of discriminatory outcomes, ensure meaningful human oversight, and maintain accuracy and transparency throughout the system’s lifecycle (EU AI Act, 2024).

Hospitals that deploy these systems have specific legal obligations too, including using AI in line with provider instructions, assigning competent human oversight, maintaining logs of system operation, and reporting serious incidents, particularly where there’s a risk of unequal outcomes across patient groups.

The AI Act doesn’t stand alone. It works alongside the Medical Device Regulation (MDR; European Union, 2017a) and In Vitro Diagnostic Regulation (IVDR; European Union, 2017b), which govern the safety of medical technologies, the GDPR (European Union, 2016), which protects the sensitive health data these systems rely on, and the emerging European Health Data Space (EHDS; European Union, 2025), designed to enable more representative datasets for future AI development. Together, they form a legal ecosystem, all ultimately grounded in the EU Charter of Fundamental Rights (2009), particularly its guarantees of non-discrimination, equality, and the right to healthcare.

All these policies translate into real responsibilities for hospitals and healthcare professionals, from understanding a tool’s limitations and knowing when to override its recommendations, to being transparent with patients about when and how these systems are used in their care.

Beyond raising awareness, the AEQUITAS project has developed a practical AI Regulatory Model that translates these legal and ethical principles into concrete steps hospitals and healthcare professionals can actually apply, turning complex EU legislation into something usable in everyday clinical practice.

References:

Beauchamp, T. L., & Childress, J. F. (2019). Principles of biomedical ethics (8th ed).

European Commission, Timeline for Implementation of the EU AI Act, Brussels: European Commission, 2024.

European Union. (2009). Charter of Fundamental Rights of the European Union. Official Journal of the European Union, C 303/1. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:12012P/TXT

European Union, Regulation (EU) 2016/679 on General Data Protection Regulation (GDPR), Brussels: EU, 2016.

European Union, Regulation (EU) 2017/745 on Medical Devices (MDR), Official Journal of the European Union, L117, Brussels: European Union, 2017.

European Union, Regulation (EU) 2017/746 on In Vitro Diagnostic Medical Devices (IVDR), Official Journal of the European Union, L117, Brussels: European Union, 2017.

European Union, Regulation (EU) 2025/327 on the European Health Data Space (EHDS), Brussels: European Union, 2025.

Floridi, L., Cowls, J., Beltrametti, M., Chatila, R., Chazerand, P., Dignum, V., Luetge, C., Madelin, R., Pagallo, U., Rossi, F., Schafer, B., Valcke, P., & Vayena, E. (2018). AI4People—An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations. Minds and Machines, 28(4), 689–707. https://doi.org/10.1007/s11023-018-9482-5
When “Neutral” Data Isn’t: Gender and Racial Bias in Biomedical AI

Bias in AI isn’t abstract. It shows up in real diagnoses, real treatment decisions, and real patients, particularly those who already face multiple layers of disadvantage. Discrimination in AI is an intersectional issue, overlapping with age, gender, race, sexual orientation, and health status, among other dimensions of identity. The patients most at risk of being harmed by biased AI tools are often those who already face the most barriers to quality care.

Decades of biased medical data

The biases we see in AI didn’t come from nowhere. They’re a direct consequence of medical data collected over decades, data shaped by systemic discrimination and structural inequalities in healthcare research and practice (Cirillo et al., 2020; Cross et al., 2024).

For much of medical history, clinical and experimental studies focused predominantly on male participants. This is what is known as the gender health gap. In 2020, only 5% of global health research funding went to women’s health research. This was split into 4% for women’s cancers and 1% for all other women-specific health conditions, with 25% of that further limited to fertility research (Nature Reviews Bioengineering, 2024).

It’s worth noting this isn’t a one-way problem: in some areas, such as depression, men are underrepresented in clinical data, largely because they’re less likely to seek care, report symptoms, or receive a diagnosis (Smith et al., 2018).

Racial bias is equally well documented

Racial bias in medicine is also extensively documented, particularly in the United States where research shows that, for instance, Black patients and other minority groups receive fewer medical procedures, lower rates of surgical intervention, and fewer referrals to specialists than white patients, regardless of clinical need (Bowser, 2001; Williams & Wyatt, 2015). When AI systems are trained on data that reflects decades of existing inequalities in healthcare access and treatment, they risk encoding and perpetuating those inequalities at a much larger scale. While much of the evidence comes from the US, the structural conditions producing racial bias, including socioeconomic inequalities, barriers to access and underrepresentation in research, are present across Europe too.

LGBTQIA+ patients face their own layer of risk

It is also important to recognise the specific situation of LGBTQIA+ individuals, who experience discrimination in healthcare and are subject to stereotypes that affect the care they receive. These social and cultural factors perpetuate discrimination and have a measurable impact on health and healthcare. Research has shown that 16% of LGBTQIA+ individuals report discrimination in healthcare encounters, and 18% avoid seeking care altogether due to fear of mistreatment (Chang et al., 2025). Another large-scale study found that in emergency department scenarios, AI recommendations for LGBTQIA+ patients included mental health interventions six to seven times more often than was clinically appropriate (Chang et al., 2025).

How this plays out in three key areas

In cardiovascular care, women’s symptoms often differ from the “classic” presentation described in medical textbooks, itself largely based on male patient data (Fatunde et al., 2025). As a result, women are offered fewer diagnostic tests, less medication, and fewer specialist referrals (Al Hamid et al., 2024). Racial bias compounds the problem: pulse oximeters, devices routinely used to measure blood oxygen saturation, have been shown to produce less accurate readings for patients with darker skin tones (Sjoding et al., 2020), a bias that then feeds directly into AI-based triage systems.

In diabetes care, racial bias is extensively documented. Studies have shown that African American patients present systematically higher A1c values than white patients with the same average blood glucose (Karter et al., 2023). If an AI system uses A1c as a proxy for glycaemic control without accounting for this, it risks producing incorrect diagnoses for Black patients.

In depression, both gender and racial bias carry significant weight, especially in tools using natural language processing for screening. Men and women tend to express psychological distress differently (Pennebaker et al., 2003), so a system trained predominantly on one gender’s language patterns may screen inaccurately for the other. Most mental health AI tools also still operate on binary gender assumptions, excluding non-binary, transgender, and gender non-conforming individuals from both the data and the populations these tools are meant to serve (Hafner et al., 2024).

Why this matters

These biases translate into delayed diagnoses, inappropriate treatments, and unequal care, every day, for real patients. Understanding these patterns is exactly what equips healthcare professionals to ask better questions, challenge assumptions, and advocate for the patients most at risk, which is precisely what the AEQUITAS training is designed to do.

References:

Chang, C.T., Srivathsa N., Bou-Khalil, C., Swaminathan, A., Lunn, M.R., Mishra, K., Koyejo, S,. Daneshjou, R. (2025). Evaluating anti-LGBTQIA+ medical bias in large language models. PLOS Digit Health 4(9): e0001001. https://doi.org/10.1371/journal.pdig.0001001

Cirillo, D., Catuara-Solarz, S., Morey, C., Guney, E., Subirats, L., Mellino, S., Gigante, A.A., Valencia, A., Rementeria, M.J., Chadha, A.S., & Mavridis, N. (2020). Sex and gender differences and biases in artificial intelligence for biomedicine and healthcare. npj Digit. Med. 3(81). https://doi.org/10.1038/s41746-020-0288-5

Cross, J. L., Choma, M. A., & Onofrey, J. A. (2024). Bias in medical AI: Implications for clinical decision-making. PLOS Digital Health, 3(11), e0000651. https://doi.org/10.1371/journal.pdig.0000651

Funding research on women’s health. (2024). Nature Reviews Bioengineering, 2, 797–798. https://doi.org/10.1038/s44222-024-00253-7

Hafner, F.S., Valdivia, A., Rocher, L. 2025. Gender Trouble in Language Models: An Empirical Audit Guided by Gender Performativity Theory. FAccT ’25: Proceedings of the 2025 ACM Conference on Fairness, Accountability, and Transparency, 1677–1695. https://doi.org/10.1145/3715275.3732112

Karter, A. J., Parker, M. M., Moffet, H. H., & Gilliam, L. K. (2023). Racial and Ethnic Differences in the Association Between Mean Glucose and Hemoglobin A1c. Diabetes Technology & Therapeutics, 25(10), 697–704. https://doi.org/10.1089/dia.2023.0153

Pennebaker, J. W., Mehl, M. R., & Niederhoffer, K. G. (2003). Psychological Aspects of Natural Language Use: Our Words, Our Selves. Annual Review of Psychology, 54(1), 547–577. https://doi.org/10.1146/annurev.psych.54.101601.145041

Sjoding, M. W., Dickson, R. P., Iwashyna, T. J., Gay, S. E., & Valley, T. S. (2020). Racial Bias in Pulse Oximetry Measurement. New England Journal of Medicine, 383(25), 2477–2478. https://doi.org/10.1056/NEJMc2029240

Smith, D.T., Mouzon, D.M., & Elliott, M. (2018). Reviewing the assumptions about men’s mental health: An exploration of the gender binary. American Journal of Men’s Health, 12(1), 78–89. https://doi.org/10.1177/1557988316630953

Bowser, R. (2001). Racial bias in medical treatment. Dick. L. Rev., 105(3), 365.

Williams, D. R., & Wyatt, R. (2015). Racial Bias in Health Care and Health: Challenges and Opportunities. JAMA, 314(6), 555. https://doi.org/10.1001/jama.2015.9260
The Hidden Ways AI Can Discriminate

When we talk about bias in AI, we are referring to “computer systems that systematically and unfairly discriminate against certain individuals or groups in favour of others” (Friedman & Nissenbaum, 1996). This is not about occasional errors, but about consistent, predictable patterns of unfair outcomes. Furthermore, biases in AI systems are complex, as they can enter the system at almost any stage, from the data it’s trained on to the way it’s ultimately used in a hospital setting. Recognising where bias originates, and how different types can reinforce each other, is the first step towards addressing it.

Three broad roots of bias

Researchers have identified three broad categories of bias that apply to all computer systems. The first is pre-existing bias, which comes from existing inequalities in society that get absorbed into the system, sometimes without anyone realising it (Friedman & Nissenbaum, 1996). The second is technical bias, which arises from the practical compromises made when translating complex human realities into clean computational form. And the third is emergent bias, which appears only after a system is deployed, as the world around it changes in ways the original design never anticipated.

Bias in the AI pipeline

These three categories give us a useful starting point. But for AI and machine learning specifically, researchers have identified seven more precise bias types that can affect these systems throughout their development and use (Suresh & Guttag, 2021):

Historical bias reflects prejudices and stereotypes already present in training data, even if the data is technically accurate. A 1990 study found that after coronary bypass surgery, male patients received pain medication significantly more often than female patients, who were instead given sedatives more frequently (Calderone, 1990). An AI trained on data like this would learn to replicate that same pattern, perpetuating the very inequality it inherited.

Representation bias happens when certain groups are underrepresented in training data. An AI trained mostly on data from one demographic will simply perform worse for everyone else. A well-known example is skin cancer detection tools that are significantly less accurate for patients with darker skin tones because the training datasets contained predominantly images from fair-skinned individuals (Guo et al., 2021).

Measurement bias creeps in when the way something is measured isn’t equally accurate or fair across groups. One striking case involved an algorithm that used healthcare costs as a proxy for how sick someone actually was, and that ended up disadvantaging Black patients, who, facing disproportionate levels of poverty, tended to spend less on healthcare than equally sick white patients (Obermeyer et al., 2019).

Aggregation bias occurs when a single, one-size-fits-all model is applied to genuinely diverse populations, ignoring the fact that the same data point can mean very different things depending on a person’s background.

Learning bias emerges from technical decisions made while building the model itself, choices that can amplify disparities already present in the data, sometimes without the developers being aware. Research has shown, for instance, that differential privacy, a technique meant to protect patient confidentiality, can end up reducing a model’s accuracy for underrepresented groups even further (Bagdasaryan & Shmatikov, 2019).

Evaluation bias happens when the datasets used to test and benchmark the model don’t reflect the real population it will actually serve, meaning a system can look accurate on paper while quietly failing specific groups of patients in practice.

Finally, deployment bias arises when a tool is used in a context very different from the one it was designed and tested for, undermining its reliability in ways that are easy to miss.

Why this matters for patients

Understanding where bias comes from is the first step to catching it, and to making sure the AI tools shaping modern medicine work fairly for everyone, not just the patients who happen to be well represented in the data.

This is exactly the kind of practical, structured understanding the AEQUITAS training equips healthcare professionals with, helping them recognise these patterns before they translate into real harm for real patients.

References:

Bagdasaryan, E., & Shmatikov, V. (2019). Differential Privacy Has Disparate Impact on Model Accuracy (arXiv:1905.12101). arXiv. https://doi.org/10.48550/arXiv.1905.12101

Calderone, K. L. (1990). The influence of gender on the frequency of pain and sedative medication administered to postoperative patients. Sex Roles, 23(11), 713–725. https://doi.org/10.1007/BF00289259

Friedman, B., & Nissenbaum, H. (1996). Bias in computer systems. ACM Transactions on Information Systems, 14(3), 330–347. https://doi.org/10.1145/230538.230561

Guo, L. N., Lee, M. S., Kassamali, B., Mita, C., & Nambudiri, V. E. (2021). Bias in, bias out: Underreporting and underrepresentation of diverse skin types in machine learning research for skin cancer detection-A scoping review. Journal of the American Academy of Dermatology, S0190-9622(21)02086-7. https://doi.org/10.1016/j.jaad.2021.06.884

Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019). Dissecting racial bias in an algorithm used to manage the health of populations. Science, 366(6464), 447–453. https://doi.org/10.1126/science.aax2342

Suresh, H., & Guttag, J. (2021). A Framework for Understanding Sources of Harm throughout the Machine Learning Life Cycle. Equity and Access in Algorithms, Mechanisms, and Optimization, EAAMO ’21, 1–9. https://doi.org/10.1145/3465416.3483305
Artificial intelligence bias in the prediction and detection of cardiovascular disease

AI can help identify people at risk of cardiovascular disease, detect it early and allow timely intervention. However, bias can arise at any step in the development, validation and use of these algorithms, causing them to perform poorly for historically marginalised groups. This article discusses the sources and consequences of AI bias in predicting and detecting cardiovascular disease, and presents strategies to reduce it.

Bias can come from training data that does not represent the population a tool is meant for, from measurements or devices that work less well for some groups, and from leaving out important factors such as social determinants of health. It can also reflect inequalities in care: because women and ethnic minority patients are more likely to have their cardiovascular disease missed or diagnosed late, algorithms trained on this data may be less accurate at detecting it in these groups. Unequal access to digital technologies can further limit how well some groups are represented in the data. The authors give several examples of AI models that performed worse for women, Black patients or older adults.

To reduce bias, the authors propose an AI health equity framework covering the whole lifecycle of an algorithm: diverse research teams, early involvement of patients, representative and balanced data, the inclusion of social determinants of health, input from clinicians, validation across different groups, transparent reporting of results and monitoring after implementation. They conclude that, without these measures, AI bias could worsen existing healthcare inequities and miss opportunities to predict and detect cardiovascular disease.

Learn more about this article here: https://doi.org/10.1038/s44325-024-00031-9

Reference:

Mihan, A., Pandey, A., & Van Spall, H.G.C. (2024). Artificial intelligence bias in the prediction and detection of cardiovascular disease. npj Cardiovascular Health, 1, 31.
AI Meets Healthcare: Understanding the Promise and the Risks

Artificial intelligence is no longer a distant, futuristic concept in medicine. It’s already here, and is becoming an everyday part of clinical practice, supporting diagnosis, clinical decision-making, treatment planning, and patient risk prediction. In fact, according to the World Health Organization, nearly three quarters of EU countries are now using AI-assisted diagnostics in some form, and that number is only expected to grow (Alowais et al., 2023; World Health Organization, 2026).

As the use of AI promises to further revolutionize the sector, what should we know before trusting it with our health?

How AI works

At its core, biomedical AI relies on three main approaches. Machine learning allows systems to learn patterns from data and make predictions without being explicitly programmed for every scenario. Deep learning, a more advanced form of machine learning, uses layered neural networks (loosely inspired by the human brain) to process complex information like medical images or clinical notes. And natural language processing (NLP) enables computers to understand and interpret human language, such as a doctor’s written notes or a patient’s description of their symptoms.

Real examples already in use

In cardiovascular care, AI tools are helping detect arrhythmias from ECG readings and estimating a patient’s individual risk of heart disease. In diabetes management, AI-powered retinal scans can catch signs of diabetic retinopathy earlier than ever before. And in mental health, natural language processing tools are being explored to help identify signs of depression in the way people write or speak, and to support patients between therapy sessions.

So what’s the catch?

Research has shown that AI applications can lead to misinterpretation and the promotion of biases and discrimination (Varona & Suarez, 2022). In essence, AI systems are only as good as the data they learn from. If that data reflects existing inequalities, for example, if it comes mostly from male patients, or predominantly from one ethnic group, the AI will inherit those biases. It will perform well for the people it “knows,” and poorly for everyone else.

Several documented cases have shown AI tools trained on biased datasets producing less accurate results for underrepresented patients, sometimes with real consequences for diagnosis and treatment.

What AEQUITAS is doing about it

This is exactly the challenge our AEQUITAS project, funded by the European Union, was designed to address. By developing training resources, a database documenting real cases of bias in biomedical AI, and a practical regulatory model for hospitals and CSOs, AEQUITAS is helping healthcare professionals and institutions across Europe use these powerful tools more safely, and more fairly, for every patient.

Understanding how AI works, and where it can go wrong, is the first step toward making sure it works for everyone.

References:

Alowais, S.A., Alghamdi, S.S., Alsuhebany, N. et al. Revolutionizing healthcare: the role of artificial intelligence in clinical practice. BMC Med Educ 23, 689 (2023). https://doi.org/10.1186/s12909-023-04698-z

Varona, D., & Suárez, J. L. (2022). Discrimination, Bias, Fairness, and Trustworthy AI. Applied Sciences, 12(12), 5826. https://doi.org/10.3390/app12125826

World Health Organization. 2026. “Artificial intelligence is reshaping health systems: state of readiness across the European Union.” World Health Organization.
The bias algorithm: how AI in healthcare exacerbates ethnic and racial disparities – a scoping review

AI health algorithms are increasingly used to predict risk and support treatment decisions, often outperforming humans in diagnostic accuracy. However, they can also underestimate risk for marginalised and underserved patients, delaying the treatment they need. This scoping review analysed 23 sources published between 2020 and 2024 to examine racial and ethnic bias in these algorithms, the role of oversight and stakeholders, and how this bias can be reduced.

A common explanation for this bias is that algorithms are trained on data in which majority populations are overrepresented, making them less reliable for others. Black, Native American, Asian and Latinx populations are often underrepresented in health datasets. Bias can also come from people: racial stereotypes absorbed during medical training can affect how clinicians assess patients, and these assessments end up in the health records used to train algorithms. The authors also describe a feedback loop: when patients from marginalised groups distrust the health system and use it less, data about them becomes even less representative, and the algorithms less accurate for them.

To address this, the review highlights the need for continuous monitoring of algorithms once they are in use, in a similar way to how medicines are monitored after approval, and for retraining or withdrawing those that create inequities. It also stresses that simply including or removing race from an algorithm does not solve the problem, and that responsibility should be shared by developers, health systems, clinicians, regulators and patients. The authors conclude that, without this oversight, AI in healthcare risks entrenching health inequities.

Learn more about this review here: https://doi.org/10.1080/13557858.2024.2422848

Reference:

Hussain, S.A., Bresnahan, M., & Zhuang, J. (2025). The bias algorithm: how AI in healthcare exacerbates ethnic and racial disparities – a scoping review. Ethnicity & Health, 30(2), 197–214.
Addressing hidden risks: Systematic review of artificial intelligence biases across racial and ethnic groups in cardiovascular diseases

AI models are increasingly used in cardiovascular medicine, but there are concerns that they may not work equally well for people of different racial and ethnic backgrounds. This systematic review examined how accurate and clinically applicable AI models for diagnosing heart disease and predicting outcomes are across racial and ethnic groups.

Of 1,704 references screened, 11 studies were included, covering diagnosis, prognosis and the analysis of heart images. Nine of them (82%) found racial or ethnic bias, while two found no differences between groups. For example, an AI model predicting heart failure from electrocardiograms was less accurate for young Black patients than for young Asian or White patients, and AI tools analysing cardiac MRI scans were more accurate for White people than for people from minority ethnic groups.

The authors conclude that AI models can perform well for some groups and less accurately for others, and that bias should be addressed at every stage, from data collection to clinical use. They recommend methods to reduce bias, standardised ways of measuring fairness, considering social determinants of health rather than race alone, and greater transparency through the sharing of code and data.

Learn more about this review here: https://doi.org/10.1016/j.ejrad.2024.111867

Reference:

Cau, R., Pisu, F., Suri, J.S., & Saba, L. (2024). Addressing hidden risks: Systematic review of artificial intelligence biases across racial and ethnic groups in cardiovascular diseases. European Journal of Radiology, 111867.
Artificial Intelligence to Promote Racial and Ethnic Cardiovascular Health Equity

AI has been seen as a tool that could help detect and reduce the racial and ethnic inequities that persist in cardiovascular care. However, AI can also carry and spread biases of its own, which means it must be approached with care. This review looks at the role of AI in cardiology in identifying and addressing these inequities, and in promoting diversity in research, as well as the concerns its use raises and strategies for using it fairly.

The authors describe how AI has already helped reduce some inequities. In one study, an AI tool identified patients with severe heart valve disease and alerted their doctors, improving follow-up care for patients of all races. Even so, Black patients still faced greater delays in receiving appropriate care than White patients, showing that AI is not a “silver bullet” for eliminating inequities. AI can also perpetuate existing biases, due to limited diversity in training data, biases within the datasets themselves, and inadequate mechanisms for detecting and monitoring bias. For example, an AI model trained mostly on data from White patients to predict heart failure from electrocardiograms performed worse for Black patients, and particularly for Black women.

To use AI in cardiovascular care in a way that addresses racial and ethnic inequities, the authors stress the need for careful design, starting with meticulous data collection and a thorough review of training data. They also call for rigorous testing and continuous monitoring of algorithms to ensure they perform equally well for all groups, as well as greater diversity in the AI workforce and the involvement of stakeholders.

Learn more about this review here: https://doi.org/10.1007/s12170-024-00745-6

Reference:

Amponsah, D., Thamman, R., Brandt, E., James, C., Spector-Bagdady, K., & Yong, C.M. (2024). Artificial Intelligence to Promote Racial and Ethnic Cardiovascular Health Equity. Current Cardiovascular Risk Reports, 18(11), 153–162.
Algorithmic Bias in AI-Based Diabetes Care: Systematic Review of Model Performance, Equity Reporting, and Physiological Label Bias

AI has transformed diabetes management, particularly through tools that predict glucose levels and support clinical decisions. However, there are concerns that these technologies may not work equally well for different racial, ethnic and socioeconomic groups. This systematic review assessed how fair AI-based glucose prediction models are, focusing on differences in performance between groups, how fairness is reported, and biases linked to physiological measures.

Of 1,243 articles initially identified, only 10 met the criteria for inclusion. Just one of them explicitly evaluated whether AI model performance differed by race. Reporting on fairness was rare: only 7% of AI diabetes studies documented ethnic or racial data, and almost none carried out fairness audits. The review also highlights known physiological differences: for the same average glucose level, African American patients tend to have higher HbA1c values than White patients. If HbA1c is used as a reference to train AI models, this difference can be built into them. Yet none of the AI studies reviewed took this into account.

The authors also note that traditional diabetes risk scores overestimate risk for White individuals and underestimate it for Black individuals, a bias that AI systems trained on these scores could inherit. They conclude that, without systematic efforts to reduce bias, AI-based diabetes technologies risk worsening existing health disparities, and call for future research to prioritise transparency, representativeness and fairness.

Learn more about this review here: https://doi.org/10.61186/ist.202502.05.04

Reference:

Alipour, M., & Alipour, A. (2025). Algorithmic Bias in AI-Based Diabetes Care: Systematic Review of Model Performance, Equity Reporting, and Physiological Label Bias. InfoScience Trends, 2(5), 33–46.
Equity in Healthcare: Analyzing Disparities in Machine Learning Predictions of Diabetic Patient Readmissions

Machine learning models can help predict which patients with diabetes are likely to be readmitted to hospital, but they need to work fairly for all patients. This study compared several of these models, including deep learning, generalised linear models, gradient boosting machines and naive Bayes, assessing not only their accuracy but also how fair their predictions were across age, gender and race.

The gradient boosting model performed best overall, with an accuracy of 82.2% and consistently low error rates for most racial and gender groups. However, across several of the models, African American and Hispanic patients had higher rates of false positives, meaning they were more likely to be wrongly predicted to be readmitted. No single model performed best for every group: some worked better for certain groups of patients and others for other groups.

The authors conclude that there is no one-size-fits-all solution, and that models need to be chosen and adjusted carefully to ensure they are both accurate and fair for all patients. According to them, fair algorithms can help reduce disparities and improve outcomes for patients with diabetes from all backgrounds.

Learn more about this study here: https://doi.org/10.1109/ICHI61247.2024.00105

Reference:

Al-Zanbouri, Z., Sharma, G., & Raza, S. (2024). Equity in Healthcare: Analyzing Disparities in Machine Learning Predictions of Diabetic Patient Readmissions. In 2024 IEEE 12th International Conference on Healthcare Informatics (ICHI) (pp. 660–669). IEEE.
Assessing the potential of GPT-4 to perpetuate racial and gender biases in health care: a model evaluation study

Large language models such as GPT-4 are seen as promising tools in health care, from automating administrative tasks to supporting clinical decisions. However, they also risk perpetuating biases and giving incorrect diagnoses, with a direct impact on patient care. This study assessed whether GPT-4 shows racial and gender biases in four possible uses: medical education, diagnostic reasoning, treatment planning and assessments of patients.

The researchers used clinical cases from medical education resources and from published research on implicit bias in health care. When asked to create clinical cases, GPT-4 did not reflect the real demographic diversity of medical conditions, and consistently produced cases that stereotyped how conditions present in different groups. For the same clinical cases, its lists of possible diagnoses were more likely to include diagnoses that stereotype certain races, ethnicities and genders.

The treatment plans and assessments it produced also varied with patients’ demographic characteristics, including recommendations for more expensive procedures and differences in how patients were perceived. The authors conclude that tools like GPT-4 urgently need thorough and transparent bias assessments for each intended use before they are introduced into clinical care.

Learn more about this study here: https://doi.org/10.1016/S2589-7500(23)00225-X

Reference:

Zack, T., Lehman, E., Suzgun, M., Rodriguez, J.A., Celi, L.A., Gichoya, J., Jurafsky, D., Szolovits, P., Bates, D.W., Abdulnour, R.-E.E., Butte, A.J., & Alsentzer, E. (2024). Assessing the potential of GPT-4 to perpetuate racial and gender biases in health care: a model evaluation study. The Lancet Digital Health, 6(1), e12–e22.
Racial disparities in continuous glucose monitoring-based 60-min glucose predictions among people with type 1 diabetes

Continuous glucose monitors, small sensors worn on the body that track blood sugar levels throughout the day, are increasingly combined with AI models that predict where glucose levels are heading. Because White people are overrepresented in medical research, these models may perform worse for other groups if they are trained mainly on White patients’ data. This study tested this using data from 205 adults and children with type 1 diabetes in the US, 101 White and 104 Black.

The researchers trained AI models to predict glucose levels 60 minutes ahead, using datasets with different proportions of White and Black participants, from 0% to 100% White. As the proportion of White participants in the training data increased, prediction errors grew slightly more for Black participants than for White participants. The difference was small but statistically significant.

When the models were then personalised to each individual, using a technique called transfer learning, this difference became smaller and was no longer statistically significant. The authors note that the differences may not be large enough to matter clinically on their own, but conclude that diverse training data is important for fair prediction models, and that personalisation may help build fairer tools.

Learn more about this study here: https://doi.org/10.1371/journal.pdig.0000918

Reference:

Thomsen, H.B., Li, L.Y., Isaksen, A.A., Lebiecka-Johansen, B., Bour, C., Fagherazzi, G., van Doorn, W.P.T.M., Varga, T.V., & Hulman, A. (2025). Racial disparities in continuous glucose monitoring-based 60-min glucose predictions among people with type 1 diabetes. PLOS Digital Health, 4(6), e0000918.
Real-world artificial intelligence-based opportunistic screening for diabetic retinopathy in endocrinology and indigenous healthcare settings in Australia

Diabetic retinopathy, damage to the retina caused by diabetes, can lead to vision loss if it is not detected early. This study tested an AI system for diabetic retinopathy screening in real-world conditions, in two endocrinology clinics in Melbourne and three Aboriginal medical services in Western Australia, and also gathered the views of patients and health professionals.

Of the 236 people screened, 203 were included in the main analysis. Overall, the system identified diabetic retinopathy requiring referral with a sensitivity of 96.9% and a specificity of 87.7%, and over 93% of participants were satisfied with the service. However, specificity was substantially lower among participants recruited from Aboriginal medical services, meaning that more people without the condition were wrongly flagged for referral. The study used clinic location as an approximation of Indigenous status.

The authors note that the AI system had been trained exclusively on images from Chinese adults, and that differences in retinal pigmentation may affect its accuracy. They also point to other possible factors, such as poorer image quality, older age and other eye conditions. These explanations were not tested in the study.

Learn more about this study here: https://doi.org/10.1038/s41598-021-94178-5

Reference:

Scheetz, J., Koca, D., McGuinness, M., Holloway, E., Tan, Z., Zhu, Z., O’Day, R., Sandhu, S., MacIsaac, R.J., Gilfillan, C., Turner, A., Keel, S., & He, M. (2021). Real-world artificial intelligence-based opportunistic screening for diabetic retinopathy in endocrinology and indigenous healthcare settings in Australia. Scientific Reports, 11, 15808.
Assessing and Mitigating Bias in Medical Artificial Intelligence: The Effects of Race and Ethnicity on a Deep Learning Model for ECG Analysis

AI models developed using data from very similar populations may not work as well for other groups, and could reflect or even worsen existing racial and ethnic disparities in health care. This study tested whether this was the case for an AI model that uses electrocardiograms (ECGs) to detect a weakened heart pumping function, known as low left ventricular ejection fraction.

The model had been developed with data from nearly 45,000 patients, 96.2% of whom were non-Hispanic White. When tested on a separate group of almost 53,000 patients, it performed consistently well across all the racial and ethnic groups analysed: non-Hispanic White, Asian, Black/African American, Hispanic/Latino and American Indian/Native Alaskan. A separate AI model was able to identify patients’ racial group from the ECG to some extent, showing that ECG characteristics do vary by race, but this did not affect the first model’s performance.

The authors conclude that, in this case, racial differences in the ECG did not affect the model’s ability to detect the condition. Even so, they recommend that the performance of all new AI tools in medicine be reported across different ethnic, racial, age and sex groups, to ensure they are used responsibly.

Learn more about this study here: https://doi.org/10.1161/CIRCEP.119.007988

Reference:

Noseworthy, P.A., Attia, Z.I., Brewer, L.C., Hayes, S.N., Yao, X., Kapa, S., Friedman, P.A., & Lopez-Jimenez, F. (2020). Assessing and Mitigating Bias in Medical Artificial Intelligence: The Effects of Race and Ethnicity on a Deep Learning Model for ECG Analysis. Circulation: Arrhythmia and Electrophysiology, 13(3), e007988.
Fairness And Bias Mitigation in AI Models for Diabetes Diagnosis: A Comparative Evaluation of Algorithmic Approaches

AI models that predict diabetes can produce biased results for certain groups of patients, and several methods have been developed to reduce this. This study compared three of these methods on the same data, a widely used dataset of 768 women aged 21 and over, assessing fairness across age groups.

Each method had different strengths. The first relied on simple, transparent decision rules, making its predictions easy to interpret, but it had no specific mechanism to ensure fairness. The second was designed to prevent the model from basing its predictions on age, and achieved perfect results on the fairness measures used while keeping accuracy at an acceptable level. The third corrected the imbalance between patients with and without diabetes in the data, which gave the highest overall accuracy but did not ensure fairness.

The authors conclude that none of the methods achieves the right balance between accuracy, fairness and interpretability on its own, and propose combining several strategies to reduce bias in AI-based diabetes diagnosis.

Learn more about this study here: https://doi.org/10.54554/jtec.2025.17.03.004

Reference:

Mohd Hamdan, M.D.H., Ab Jabal, M.F., Abdul Rahman, S., & Kapi, A.Y. (2025). Fairness And Bias Mitigation in AI Models for Diabetes Diagnosis: A Comparative Evaluation of Algorithmic Approaches. Journal of Telecommunication, Electronic and Computer Engineering, 17(3), 27–34.
Mitigating bias in deep learning for diagnosis of coronary artery disease from myocardial perfusion SPECT images

AI can accurately diagnose coronary artery disease from myocardial perfusion imaging, a type of scan that shows blood flow to the heart muscle. However, these models are often trained on high-risk patients who were referred for invasive tests, which can lead them to overestimate the probability of disease in other patients. This study tested ways of training AI so that the probabilities it estimates match the real probability of disease more closely.

Deep learning models were trained on 828 patients from three sites and then tested on a separate group of 511 patients, including 179 with a low likelihood of disease. The best results came from a model trained with additional examples of patients without coronary artery disease and with little or no sign of reduced blood flow on the scan. It produced more accurate probability estimates than the model trained without these additional cases, and the improvement was most pronounced in women. It was also better at identifying disease than standard measures taken from the scans.

The authors conclude that adding more low-risk patients during training can improve how accurately AI estimates the probability of disease. This matters especially for lower-risk patients and for women, for whom overestimated probabilities could affect decisions about further tests and treatment.

Learn more about this study here: https://doi.org/10.1007/s00259-022-05972-w

Reference:

Miller, R.J.H., Singh, A., Otaki, Y., Tamarappoo, B.K., Kavanagh, P., Parekh, T., Hu, L.-H., Gransar, H., Sharir, T., Einstein, A.J., Fish, M.B., Ruddy, T.D., Kaufmann, P.A., Sinusas, A.J., Miller, E.J., Bateman, T.M., Dorbala, S., Di Carli, M.F., Liang, J.X., … Slomka, P.J. (2023). Mitigating bias in deep learning for diagnosis of coronary artery disease from myocardial perfusion SPECT images. European Journal of Nuclear Medicine and Molecular Imaging, 50(2), 387–397.
Evaluating and mitigating bias in machine learning models for cardiovascular disease prediction

Machine learning models are increasingly used to estimate a person’s risk of cardiovascular disease, but bias can be introduced when health data is collected and prepared, affecting how well these models work for certain groups. This study used electronic health records from 109,490 people treated at Vanderbilt University Medical Center, in the US, to test whether several machine learning models performed equally well across race and gender. It also compared them with a widely used risk equation from the American Heart Association that does not use machine learning.

Most of the machine learning models were less biased than the American Heart Association equation. Differences between racial groups were small, but differences between women and men were considerably larger. According to the authors, both the risk equation and the machine learning models were biased against women.

The researchers also tested three ways of reducing this bias. Removing race and gender from the data did not significantly reduce bias in most models, and balancing the number of people from each group in the training data did not consistently help either. Balancing the proportion of people with cardiovascular disease in each group reduced the gender bias, but not the racial bias, and slightly reduced accuracy in many cases. The authors conclude that gender disparities in cardiovascular risk prediction need to be investigated and corrected.

Learn more about this study here: https://doi.org/10.1016/j.jbi.2023.104294

Reference:

Li, F., Wu, P., Ong, H.H., Peterson, J.F., Wei, W.-Q., & Zhao, J. (2023). Evaluating and mitigating bias in machine learning models for cardiovascular disease prediction. Journal of Biomedical Informatics, 138, 104294.
Race, Sex, and Age Disparities in the Performance of ECG Deep Learning Models Predicting Heart Failure

AI models that analyse electrocardiograms (ECGs) could help identify people at high risk of heart failure early, and potentially reduce racial disparities in heart failure outcomes. But whether these models perform equally well for different groups of patients has been little studied. This study used ECGs from 326,518 patient encounters at Stanford Hospital to evaluate a model trained to predict heart failure within five years, comparing its performance by race, ethnicity, age and sex.

Overall, there were no significant differences in performance between racial groups, and performance declined with age. However, among patients aged 0 to 40, the model performed significantly worse for Black patients than for any other racial group, with the largest differences among young Black women.

The researchers tested several ways of reducing these disparities: including race, ethnicity, sex and age in the model, training separate models for each racial group, and training it on data with equal racial representation. None of these improved the disparities. Adjusting the model’s decision thresholds individually by race, age and sex did lead to substantial improvements. The authors warn that machine learning tools for heart failure risk perpetuating existing disparities, and suggest these individualised thresholds as one way to reduce them.

Learn more about this study here: https://doi.org/10.1161/CIRCHEARTFAILURE.123.010879

Reference:

Kaur, D., Hughes, J.W., Rogers, A.J., Kang, G., Narayan, S.M., Ashley, E.A., & Perez, M.V. (2024). Race, Sex, and Age Disparities in the Performance of ECG Deep Learning Models Predicting Heart Failure. Circulation: Heart Failure, 17(1), e010879.
Predictive Accuracy of Stroke Risk Prediction Models Across Black and White Race, Sex, and Age Groups

Stroke is a leading cause of death and serious long-term disability in the US, with a particularly high risk among Black individuals. Risk prediction tools help identify who would benefit most from prevention, but they need to work equally well for everyone. This study compared several of these tools, including two stroke-specific models, a general cardiovascular risk model and two new machine learning models, using data from 62,482 Black and White adults aged 45 and over from four large US studies.

Overall, the models performed similarly, and the new machine learning models did not improve their accuracy. However, all of them were worse at distinguishing between people who would and would not have a stroke when applied to Black individuals. Among women, the models’ accuracy score was 0.76 for White women and 0.69 for Black women. Among men, it ranged from 0.71 to 0.72 for White men and from 0.64 to 0.66 for Black men. The models also performed better in adults under 60 than in older adults.

The authors conclude that current approaches are not enough to close this gap, and that the range of risk factors considered and the modelling techniques used need to be improved to address these racial disparities.

Learn more about this study here: https://doi.org/10.1001/jama.2022.24683

Reference:

Hong, C., Pencina, M.J., Wojdyla, D.M., Hall, J.L., Judd, S.E., Cary, M., Engelhard, M.M., Berchuck, S., Xian, Y., D’Agostino, R., Howard, G., Kissela, B., & Henao, R. (2023). Predictive Accuracy of Stroke Risk Prediction Models Across Black and White Race, Sex, and Age Groups. JAMA, 329(4), 306–317.
Confounders mediate AI prediction of demographics in medical imaging

Previous research has suggested that AI can identify a patient’s age, sex and even race from medical images. This study tested whether this holds for echocardiograms, ultrasound videos of the heart, using more than 530,000 videos from two US medical centres, Cedars-Sinai and Stanford.

The models predicted sex well and age with an average error of about 9 years, and these results held across both centres. Race, however, could not be reliably predicted. When the researchers changed the balance of age or sex in the data used to train the race model, its performance varied widely, while changing the racial composition of the data had little effect on the predictions of sex and age.

The authors conclude that much of the model’s apparent ability to predict race comes from “shortcuts” through other features that differ between racial groups, rather than from race itself. They warn that treating AI predictions of race as evidence of biological differences risks reinforcing a false idea that has contributed to medical errors, since differences in health between racial groups often reflect social factors and structural racism. They call for further research into which imaging features carry demographic information, and into the risks of bias this creates in medical AI.

Learn more about this study here: https://doi.org/10.1038/s41746-022-00720-8

Reference:

Duffy, G., Clarke, S.L., Christensen, M., He, B., Yuan, N., Cheng, S., & Ouyang, D. (2022). Confounders mediate AI prediction of demographics in medical imaging. npj Digital Medicine, 5, 188.
Addressing Artificial Intelligence Bias in Retinal Diagnostics

Diabetic retinopathy, damage to the retina caused by diabetes, is one of the conditions in which AI is used to support diagnosis by analysing images of the back of the eye. This study looked at what happens when the data used to train these systems is unbalanced, and tested a way of reducing the resulting bias.

The researchers used a public dataset of retinal images to which clinicians had added skin-tone labels. To recreate a situation of imbalance, they removed from the training data the images of people with darker skin who had diabetic retinopathy requiring referral to a specialist, while keeping these images in the test data. The standard system was 73.0% accurate for people with lighter skin and 60.5% for people with darker skin, a gap of 12.5 percentage points.

They then trained the system with additional synthetic retinal images created by generative AI models. Accuracy became 72.0% for people with lighter skin and 71.5% for people with darker skin, reducing the gap to 0.5 points. The authors conclude that unbalanced training data can lead to unequal accuracy across groups, and that synthetic images may help reduce this bias in the diagnosis of diabetic retinopathy and possibly in other AI systems used in eye care.

Learn more about this study here: https://doi.org/10.1167/tvst.10.2.13

Reference:

Burlina, P., Joshi, N., Paul, W., Pacheco, K.D., & Bressler, N.M. (2021). Addressing Artificial Intelligence Bias in Retinal Diagnostics. Translational Vision Science & Technology, 10(2), 13.
Age, sex and race bias in automated arrhythmia detectors

Algorithms that read electrocardiograms (ECGs) automatically are increasingly used to help identify heart conditions, but few studies have examined whether they perform equally well for different groups of patients. This study compared 56 open-source algorithms, developed by participants in the 2021 PhysioNet Challenge, on more than 130,000 ECGs, grouping the results by sex, age and race.

Several performance measures were similar across groups, but others showed clear differences. Accuracy was significantly lower for Black patients and higher for Asian patients, with a gap of 39% on the test set. Accuracy also fell with age, by around 12% per decade, and results for male and female patients differed depending on the measure used.

The researchers then retrained one of the best-performing models with an additional constraint designed to reduce performance differences across sex, race and age. This led to a small loss in overall performance but a significant reduction in bias. The authors conclude that bias depends on the model’s design, the population it is trained on and how its performance is measured, and that all of these should be closely examined.

Learn more about this study here: https://doi.org/10.1016/j.jelectrocard.2022.07.007

Reference:

Perez Alday, E.A., Rad, A.B., Reyna, M.A., Sadr, N., Gu, A., Li, Q., Dumitru, M., Xue, J., Albert, D., Sameni, R., & Clifford, G.D. (2022). Age, sex and race bias in automated arrhythmia detectors. Journal of Electrocardiology, 74, 5–9.
Artificial intelligence-enhanced electrocardiography for accurate diagnosis and management of cardiovascular diseases

Electrocardiography (ECG) enhanced by artificial intelligence is emerging as a promising tool for diagnosing and managing cardiovascular disease. This review explains how AI can detect subtle patterns in the ECG that human interpreters may miss, helping to identify problems such as arrhythmias and heart failure more quickly and consistently.

However, the authors also highlight important limitations, including systematic bias. When training datasets do not adequately represent different ages, sexes and racial groups, models can perform worse for those who are underrepresented, since ECG patterns vary with age, physiological differences between the sexes and genetic factors. To make these tools safe and fair in clinical practice, the review calls for more diverse datasets, bias mitigation strategies and continuous validation across populations.

Learn more about this review here: https://doi.org/10.1016/j.jelectrocard.2024.01.006

Reference:

Muzammil, M.A., Javid, S., Afridi, A.K., Siddineni, R., Shahabi, M., Haseeb, M., Fariha, F., Kumar, S., Zaveri, S., & Nashwan, A.J. (2024). Artificial intelligence-enhanced electrocardiography for accurate diagnosis and management of cardiovascular diseases. Journal of Electrocardiology, 83, 30–4
Cardiovascular Disease Screening in Women: Leveraging Artificial Intelligence and Digital Tools

Cardiovascular disease remains the leading cause of death in women, and there is growing evidence of sex- and gender-based differences in how it develops and progresses. This review provides an overview of the main opportunities for cardiovascular screening in women and of the potential role of artificial intelligence and digital tools.

The authors note that healthcare providers, guideline committees and policymakers are not yet well equipped to collect data on women, interpret it, or use it to design targeted interventions. The review includes a section on bias in AI and looks at different stages of a woman’s life, from youth and pregnancy to menopause and older age, as well as the implications of AI for clinical guidelines and public health.

Learn more about this review here: https://doi.org/10.1161/CIRCRESAHA.121.319876

References:

Adedinsewo, D.A., Pollak, A.W., Phillips, S.D., Smith, T.L., Svatikova, A., Hayes, S.N., Mulvagh, S.L., Norris, C., Roger, V.L., Noseworthy, P.A., Yao, X., & Carter, R.E. (2022). Cardiovascular Disease Screening in Women: Leveraging Artificial Intelligence and Digital Tools. Circulation Research, 130(4), 673–690.