Research Platform Published Research: Annals of Vascular Surgery 14 ESVS Guidelines Unified

Research over published vascular surgery guidelines

ESVS Guideline Research Tool

ClinicalGuidelines.io is a retrieval-augmented clinical decision support tool over 14 ESVS guidelines.

For research and evaluation only. Not a medical device and not for clinical use. Do not enter real patient-identifiable information. Responses are generated from published ESVS guidelines and have not been validated for clinical accuracy.

14

Published ESVS guidelines form the system's retrieval scope.

Router

Routes requests among the included guidelines.

Context Gate

Requests missing detail in predefined scenarios; coverage is incomplete.

Sources

Retrieval uses narrative text and structured recommendations.

Query Handling

The system handles guideline questions and requests missing clinical detail in predefined scenarios. This check does not cover all presentations.

Knowledge Questions

Guideline questions proceed to retrieval from the included sources.

Specified Research Cases

Cases with the parameters requested by the system proceed to retrieval.

Missing Clinical Detail

The Context Gate requests missing clinical detail in predefined scenarios; it does not cover all presentations.

Questions Outside Scope

The system will decline questions outside the 14 ESVS guidelines.

Published Research

The chart below relates to the earlier single-guideline application published in Annals of Vascular Surgery.

The earlier study does not establish clinical accuracy for the 14-guideline system. Its answers have not been assessed for clinical acceptability by expert reviewers; that evaluation is planned.

14

ESVS Guidelines

The system's retrieval scope

Dual

Retrieval Channels

Narrative text and structured recommendations

Context Gate

Clarification Prompts

Requests missing detail in predefined scenarios

Retrieval Architecture

The pipeline incorporates query interpretation, guideline routing, clarification prompts, and retrieval from narrative text and structured recommendations.

01

Pre-retrieval interpretation

The incoming query is classified as a guideline question or a case for which clinical detail may be requested.

02

Guideline Router Service

The router selects among the 14 included ESVS guidelines.

03

Context Gate

The system requests missing clinical detail before advising, but this check does not cover all presentations.

04

Dual-channel retrieval

Retrieval uses both narrative guideline text and structured recommendations.

05

Response generation

Responses are generated from published ESVS guidelines and have not been validated for clinical accuracy.

Known limitations.

The system is constrained to 14 ESVS guidelines and will decline questions outside them. It requests missing clinical detail before advising, but this check does not cover all presentations. Its answers have not been assessed for clinical acceptability by expert reviewers; that evaluation is planned. PHI de-identification has not been tested on real data.

Planned Evaluation

Assessment of the answers for clinical acceptability by expert reviewers is planned.

Research plans

Hellenic Society of Vascular and Endovascular Surgery

Prospective multicenter evaluation in conjunction with HSVES is planned.

The system's answers have not been assessed for clinical acceptability by expert reviewers.

Publications

Published studies from the research group in Annals of Vascular Surgery.

1

Alexiou VG, Sumpio BE, Vassiliou A, Kakkos SK, Geroulakos G.

Artificial Intelligence in Diagnosing and Managing Vascular Surgery Patients: An Experimental Study Using the GPT-4 Model

Ann Vasc Surg. 2025 Feb;111:260-267. doi: 10.1016/j.avsg.2024.11.014. Epub 2024 Nov 24. PMID: 39586530.

An earlier experimental study using GPT-4.

Presentations

Conference presentations from the research group. Titles are reproduced as presented and do not establish clinical accuracy or regulatory approval of the system.

Oral Presentation 2025

39th ESVS Annual Meeting

Istanbul, Türkiye

23-26 September 2025

Session window
Tuesday 23 September, 13:00 - 17:30
Reference
AS-ESVS-2025-00397
Abstract
An Artificial Intelligence companion app that provides expert-level interpretation of clinical practice guidelines
Conference Presentation 2025

SITE Barcelona 2025

Barcelona, Spain

15-17 October 2025

Presentation of the ClinicalGuidelines.io concept.

Congress Presentation 2026

25th Panhellenic Congress of Vascular & Endovascular Surgery - Angiology

Larissa Imperial, Larissa, Greece

19-21 March 2026

Session
Friday, 20 March 2026, 09:00 - 10:00
Session 3: Innovation, Technology, and Entrepreneurship in Vascular Surgery
Talk
No Hallucinations, Just Guidelines: CE-Marking an AI Companion as a Medical Device

Scientific Leadership

A multidisciplinary leadership group spanning vascular surgery, academic medicine, and translational clinical AI.

Vangelis Alexiou, MD, PhD

Assistant Professor of Vascular Surgery

University of Patras, Greece

Areti Vassiliou, MD, MSc

Vascular Fellow

Attikon University Hospital, Athens

Prof. Bauer E. Sumpio, MD, PhD

Professor of Surgery

Yale University School of Medicine, USA

Prof. George Geroulakos, MD, PhD

Professor of Vascular Surgery

National and Kapodistrian University of Athens

Prof. Stavros K. Kakkos, MD, PhD

Professor of Vascular Surgery

University Hospital of Patras, Greece

Request Research Access

Request access for research and evaluation or register interest in the planned expert assessment.

For research and evaluation only. Not a medical device and not for clinical use. Do not enter real patient-identifiable information. Responses are generated from published ESVS guidelines and have not been validated for clinical accuracy.