185: The Patient's Voice: AI and Digital Pathology in Cancer Care

Send us Fan Mail

Source Material: This AI journal Club episode is based on the original article, "The patient matters: a roundtable discussion on pathology in the era of digitization and AI," authored by Frederik Deman, Heleen Lauwers, Glenn Broeckx, Roberto Salgado, and Amelie Dendooven (Virchows Archiv, 2026).

In this episode, we dive deep into a critical yet often overlooked aspect of medical technology: the patient's perspective on the rapid integration of Digital Pathology (DP) and Artificial Intelligence (AI) in oncological diagnostics. While AI and digital tools promise faster and more precise cancer diagnoses, patient viewpoints have historically been absent from the conversation. We unpack the insights from a dedicated roundtable discussion involving six Flemish cancer-patient advocates, revealing their hopes, practical expectations, and surprising opinions on AI in the lab.

Key Takeaways & Topics Covered:

• Breaking Down Physical Barriers: We explore how patients view digital pathology as a way to easily share high-resolution tissue images across institutions, allowing subspecialized pathologists to collaborate and provide rapid second opinions without the delays of shipping physical slides.

• Trust, Accuracy, and Human Oversight: Patients exhibit a high level of trust in AI to speed up workflows and extract more precise data from biopsies. However, they emphasize that trust is contingent on algorithms being trained on diverse, high-quality data and pathologists retaining ultimate oversight.

• The "Black Box" Debate: Do patients need to know exactly how AI makes its decisions? Surprisingly, no. Patients prioritize clinical validity and accuracy over full algorithmic transparency. If a tool is proven safe and reproducible, a lack of "explainability" is not seen as a barrier to its use, though they encourage scientists to keep studying how these models work.

• Redefining Privacy Fears: The advocates were largely unconcerned with their pseudonymized data being transferred to cloud-based AI platforms. Their overriding privacy fear was not hacking, but rather the potential for health data to be accessed by insurance providers, which could lead to denied coverage or discrimination.

• AI as a Patient Empowerment Tool: Looking ahead, patients strongly desire future AI applications that can translate jargon-heavy pathology reports into lay-friendly language. They also envision AI generating customized questions about prognostic implications or follow-up tests, allowing patients to have more active, focused consultations with their doctors.

• The Cost of Innovation: We discuss the financial tension of AI in healthcare. While some patients are willing to pay a modest premium for faster AI-assisted diagnoses, there is a strong consensus that linking advanced tech to out-of-pocket payments could exacerbate healthcare inequality.

• Clashing with the EU AI Act: We conclude with a fascinating contrast: the pragmatic stance of patients—who accept non-explainable AI as long as it works safely—appears to misalign with the strict transparency and interpretability requirements mandated by the upcoming EU AI Act.

Support the show

Get the "Digital Pathology 101" FREE E-book and join us!