The patient matters: a roundtable discussion on pathology in the era of digitization and AI

Abstract

Digital pathology (DP) and artificial intelligence (AI) promise faster, more accurate cancer diagnostics, yet patient views remain undocumented. We explored perspectives of patient representatives on DP and AI implementation. A two-hour moderated roundtable with six Flemish cancer-patient advocates was recorded, transcribed and analyzed using reflexive thematic analysis. Participants anticipated improved accuracy, shorter turnaround times and stronger inter-laboratory collaboration. Trust in AI was high when algorithms were trained on diverse datasets and pathologists retained final responsibility. Clinical validity outweighed full algorithmic transparency, though ongoing explainability research was encouraged. Explicit mention of AI in reports was considered unnecessary if quality assurance was demonstrable. Privacy worries focused on potential insurer misuse rather than pseudonymized cloud transfer. Representatives requested future tools that translate technical reports into lay language and suggested questions to support shared decision-making. Patient representatives were generally supportive of the introduction of AI in pathology, provided that algorithms are clinically validated, trained on representative datasets, and deployed under clear professional oversight. Their comments specifically highlight expectations regarding human-AI collaboration, data governance, auditability, and communication about AI use. These AI-focused insights can help laboratories, vendors, and regulators align development and implementation with patient priorities.

Ähnliche Arbeiten