Quality Control of Histological Preparations in Digital Pathology: Development of an Evaluation Score for Routine Technical Prepara-tion According to the Lean Six-Sigma Methodology

Abstract

The technical quality of histological preparations is a fundamental prerequisite for ensuring diagnostic accuracy in pathology and is a prerequisite for the implementation of digital pathology and artificial intelligence applications. However, standardized tools for the systematic and objective assessment of histological quality are lacking.

An integrated score based on predefined criteria and a Six Sigma approach was developed and applied, with the aim of measuring technical defects throughout the various phases of the histological process (embedding, cutting, staining, coverslipping, traceability). The data were analyzed on various sample types (biopsies, bone marrow, skin), identifying both critical areas and process robustness points.

The analysis revealed an overall mean score of 2.8 Sigma, just below the acceptability threshold (≥3 Sigma). The main non-conformities involved cutting defects (torn, bent, or incomplete sections) and inadequate thickness, with a greater impact on bone marrow biopsies. Further defects were detected during the slide covering phase, especially in skin samples. Conversely, traceability processes (absence of QR code errors) and the quality of staining and vitrified surfaces were found to be robust.

The proposed score appears to be an effective tool for objective and reproducible assessment of histological quality, useful for both continuous monitoring and decision-making support in laboratories. The identified critical issues require targeted corrective actions (instrument calibration, protocol standardization, operator training). Looking ahead, integrating this model into digital systems and multicenter networks could foster the creation of a standardized digital pathology ecosystem, progressively improving the diagnostic quality and reliability of Artificial Intelligence systems.