Toward Auditable Neuro-Symbolic Reasoning in Pathology: SQL as an Explicit Trace of Evidence

TL;DR

This paper presents an SQL-based framework for transparent and auditable pathology image analysis, enabling explicit reasoning and evidence tracing from cellular features to diagnostic decisions.

Contribution

It introduces an SQL-centered agentic system that extracts features, composes queries, and evaluates findings against pathology knowledge for improved interpretability.

Findings

Enhanced interpretability and traceability in pathology diagnosis.

SQL traces link cellular features to conclusions effectively.

Improved decision justification in visual question answering datasets.

Abstract

Automated pathology image analysis is central to clinical diagnosis, but clinicians still ask which slide features drive a model's decision and why. Vision-language models can produce natural language explanations, but these are often correlational and lack verifiable evidence. In this paper, we introduce an SQL-centered agentic framework that enables both feature measurement and reasoning to be auditable. Specifically, after extracting human-interpretable cellular features, Feature Reasoning Agents compose and execute SQL queries over feature tables to aggregate visual evidence into quantitative findings. A Knowledge Comparison Agent then evaluates these findings against established pathological knowledge, mirroring how pathologists justify diagnoses from measurable observations. Extensive experiments evaluated on two pathology visual question answering datasets demonstrate our method improves interpretability and decision traceability while producing executable SQL traces that link cellular measurements to diagnostic conclusions.