Belgian Study: Skin Cancer Detection App Inaccurate

A new study from Ghent University and Ghent University Hospital in Belgium raises questions about the reliability of skin cancer detection apps, specifically SkinVision. The research, published in late February 2026, indicates the app misses a significant proportion of actual skin cancers – up to one in four – and generates false alarms in nearly one in eight cases of benign skin lesions. This finding is particularly relevant as SkinVision is offered by several health insurance funds in Belgium, making it a readily accessible tool for many.

How SkinVision Works and Why Accuracy Matters

SkinVision operates by allowing users to photograph a skin lesion using their smartphone. Artificial intelligence then analyzes the image and provides an assessment of whether the spot is suspicious or not. The app isn’t intended to provide a diagnosis, but rather to offer initial guidance, potentially prompting users to seek professional medical evaluation. The increasing popularity of such apps highlights a growing trend toward self-assessment of health concerns, driven by convenience and accessibility. However, the accuracy of these tools is paramount, as both missed cancers and unnecessary anxiety from false positives can have significant consequences.

The study, detailed in reporting by VRT NWS, involved a prospective evaluation of the app’s performance. Researchers assessed its ability to correctly identify both cancerous and non-cancerous skin lesions in a real-world setting. The results underscore the limitations of relying solely on smartphone apps for skin cancer screening.

Study Details and Ongoing Research

The research builds on a broader doctoral project at Ghent University, titled “AI-based tools for skin cancer detection: are we ready for it?” As outlined on the Research Explorer, this project aims to evaluate the accuracy and safety of AI-driven systems for skin cancer diagnosis at both the consumer and clinical levels. The current study focused on the consumer-facing app, SkinVision, and its performance among general users. The project is scheduled to run until September 2026, suggesting further analysis and refinement of these tools are anticipated.

A related study, as noted in the Research Explorer project description, acknowledges the availability of numerous skin cancer detection apps but emphasizes the ongoing concerns regarding their accuracy in realistic conditions. The research team is specifically investigating factors that influence app performance and user acceptance.

Understanding False Positives and Missed Cancers

A false positive result – where the app incorrectly identifies a benign lesion as potentially cancerous – can lead to unnecessary anxiety, follow-up appointments, and potentially even biopsies. While not physically harmful, these interventions carry emotional and financial costs. Conversely, a missed cancer – where the app fails to identify a malignant lesion – can delay diagnosis and treatment, potentially impacting patient outcomes. The study’s finding that up to 25% of skin cancers were missed is a particularly concerning aspect of the research.

It’s important to note that SkinVision has responded to the study, stating that the research is outdated and that the accuracy of their AI algorithms improves with each iteration. The company emphasizes that the app is intended as a supportive tool, not a substitute for professional medical diagnosis. However, the study’s findings raise questions about the extent to which these improvements translate into real-world accuracy for users.

The Role of AI in Skin Cancer Screening: A Broader Perspective

The use of artificial intelligence in healthcare is rapidly expanding, with applications ranging from image analysis to drug discovery. In the context of skin cancer screening, AI-powered tools offer the potential to increase access to early detection, particularly in areas with limited access to dermatologists. However, the accuracy and reliability of these tools are critical considerations. The study from Ghent University highlights the need for rigorous, independent evaluation of these technologies before widespread adoption.

Researchers are also exploring the potential for AI to assist dermatologists in risk assessment and diagnosis. The ResearchPortal.be details a project investigating whether apps like SkinVision can help prioritize lesions for urgent medical attention. This approach aims to leverage AI to improve the efficiency of the diagnostic process, rather than replacing the role of the clinician.

What This Means for You: Guidance and Next Steps

If you have used SkinVision or a similar app and received a concerning result, it is crucial to consult with a qualified healthcare professional – a dermatologist or your primary care physician – for a proper evaluation. Do not rely solely on the app’s assessment for diagnosis or treatment decisions. Conversely, a negative result from the app does not guarantee that a lesion is benign; regular self-exams and professional skin checks remain essential.

The findings from Ghent University underscore the importance of ongoing research and independent evaluation of AI-powered healthcare tools. As these technologies continue to evolve, it will be crucial to ensure that they are accurate, reliable, and used appropriately to improve patient care. Further studies are needed to assess the long-term impact of these apps on skin cancer detection rates and patient outcomes.

Staying Informed: Maintain abreast of official guidance from reputable health organizations regarding skin cancer screening and prevention. The World Health Organization (WHO) and national health authorities provide valuable resources and recommendations.