• There are no suggestions because the search field is empty.

Digital pathology deployment in Veneto: 8 steps to success

What is required for pathology laboratories to go fully digital? Learn how the broadest public healthcare provider in the Veneto region did it.
Written by Aiforia

Azienda Zero is the broadest public healthcare provider in Veneto’s North Italian region, and its healthcare system serves almost 5 million people. In the region's 12 hospitals, the routine workflow of the pathology laboratories accounts for about 240,000 slides per month and nearly 3 million slides per year.

This volume, albeit not the only reason, makes the laboratories suitable for adopting a shared digital pathology system. The digitalization project aims to create a safe, standardized process to improve patients’ clinical care. 

This article covers the eight components of Veneto’s digital pathology transformation project. It is based on Eccher et al.'s (2024) recent publication, “Digital pathology structure and deployment in Veneto: a proof-of-concept study”.

8 components in the digital pathology workflow

  1. Image‑acquisition system
  2. Workstation and viewing software
  3. Laboratory information system
  4. Storage
  5. Artificial intelligence
  6. Specialist networks
  7. System tracking, quality checks, and archiving of tissue specimens
  8. Validation

Let's review these one by one.

1. Image‑acquisition system

The first step of going digital is the implementation of scanners. The technical characteristics required heavily depend on the needs of the laboratories. In Veneto, four scanner categories were selected to cater to all the different use cases:

  • High throughput scanners for routine diagnosis in high-volume institutions
  • Medium throughput scanners for the smaller community hospitals 
  • Low-throughput scanners for fluorescent microscopy and on-site intraoperative examination
  • Scanners for the specific needs of cytopathology, where scanning of multiple focal planes typically results in longer scanning times and larger digital files 


2. Workstation and viewing software

In the Veneto project, 230 workstations were used to let pathologists, residents, students, other professionals, and meeting attendants visualize and navigate whole slide images. Every station included two displays: one for showing histological images and the other for viewing data. The displays were medical-grade US Food and Drug Administration-approved full HD monitors (resolution 1920 × 1080 pixels), at least 27 inches in size.

Another related component was the viewing software pathologists can use to analyze displayed whole slide images (WSIs). The viewing software allows the users to visualize virtual slides and navigate WSIs with various annotation functions, such as drawing regions of interest, zooming in and out, rotating, and measuring.

3. Laboratory information system

The same web-based software was chosen as a regional laboratory information system (LIS) for all the laboratories in the Veneto region. The goal was to unify the pathology workflow for all laboratories in the network.

4. Storage

According to the guidelines created by the UK’s Royal College of Pathologists, histology slides and smears should be preserved for 10 years and blocks for the duration of a patient’s entire life. When planning a storage solution for digital files, the following factors should be considered: the number and size of the files, the number of users who will access the files, the interface between users, the level of data protection and security, and the budget.

In the Veneto project, the digital storage process of WSIs consisted of two archives: a short-term local archive for initial local management at the laboratories and a cloud-based central archive that allowed easy and fast case identification and retrieval.

5. Artificial intelligence

AI implementation was a key component in the Veneto project, helping to enhance diagnostic accuracy and reduce subjective interpretations and turn-around times. The project team chose to fully embrace Aiforia’s AI-based deep-learning tools, including:

  • Aiforia® Prostate Cancer Gleason Grade Groups – which supports the detection of adenocarcinoma and assists Gleason grade grouping from whole slide images (WSI).
  • Aiforia® Breast Cancer Ki67, ER, and PR – which support the detection of invasive carcinoma and scoring of Ki67, ER, and PR positive and negative tumor cells from whole slide images (WSI) or from selected image areas of cancerous breast tissue.
  • Aiforia® Lung Cancer PD-L1 – which supports the detection of invasive carcinoma areas in patient samples and the scoring of PD-L1 negative and positive tumor cells in non-small cell lung (NSCLC) cancer cases. 

All of the above-mentioned AI models and the matching Aiforia® Clinical Suite Viewer are CE-IVD marked for diagnostic use in EU and EEA countries and for Research Use Only (RUO) and Performance Studies Only (PSO) in all other market areas. 

Check what Professor Angelo Paolo Dei Tos from the University of Padova has to say on the value of digital pathology and AI ↓


Read more about Aiforia’s clinical solutions here → 


6. Specialist networks

The Veneto project aims to improve the management of complex nephropathology cases by structuring reporting workstations within nephrology departments similar to those in pathology laboratories. This will allow the entire regional nephrology network to be connected, facilitate the management of complex cases, enable process quality controls, standardize reporting, and ultimately enhance the therapeutic management of patients.

7. System tracking, quality checks, and archiving of tissue specimens

To maintain accuracy and consistency throughout the workflow, the pathology laboratories involved in the project incorporated numerous quality control checkpoints. 

Each specimen that arrives at the laboratory is labeled with a unique barcode that can be used to track it throughout its lifecycle. The barcode-scanning equipment was integrated with the local laboratory information system.

8. Validation

The College of American Pathologists (CAP) has set guidelines for validating WSI systems for diagnostic purposes in pathology. The goal is to ensure that the WSIs will perform as expected for their intended use and environment before using them for patient care. 

According to the CAP guidelines, a validation set must include at least 60 cases for one application or use case reflecting the spectrum and complexity of specimen types. Multiple pathologists per center, experienced in digital pathology, will enhance the validation process by randomly evaluating selected WSIs previously assessed by conventional light microscopy (LM) after the washout period (minimum two weeks). The target for the diagnostic concordance threshold between LM and WSI is 95%. Although all discrepancies between WSI and glass-slide diagnoses identified during validation need to be resolved, achieving this concordance threshold is crucial.

Benefits of digital pathology

This digital transformation project in the Veneto region is a significant milestone for the further widespread adoption of digital pathology and AI in Italy. 

The project’s estimated benefits include: 

1) Cost savings 

Several studies have estimated cost savings of more than $250,000 per institution per year as a result of implementing digital pathology. This is also expected in the Veneto region due to decreased operational and asset costs. 

2) Improved exchange of information 

The wide deployment of digital pathology, WSI, and AI will substantially improve the information flow in the hospital network. This, in turn, will enhance the standardization and reproducibility of pathology reports. Physicians can also obtain remote second opinions to enhance the general clinical management of patients.

3) Education opportunities

An extensive archive of WSIs will create substantial pathology repositories that help residents, students, and young pathologists widen their knowledge. Easily accessible digital data will also speed up the adoption of AI-guided diagnoses, as it can be used for external validation of already available algorithms and as the basis for training new ones. 

Read also: