RAM-H1200: A Unified Evaluation and Dataset on Hand Radiographs for Rheumatoid Arthritis

2026-05-06

AI Summary

This study addresses the lack of publicly available datasets supporting unified, multi-level analysis of rheumatoid arthritis (RA) in hand radiographs, particularly the absence of full-hand coverage, fine-grained bone erosion annotations, and integration with clinical scoring systems. To bridge this gap, the authors introduce a public benchmark dataset comprising 1,200 multicenter hand X-ray images, which for the first time jointly provides full-hand bone instance segmentation, pixel-level bone erosion masks, joint regions defined by the Sharp–van der Heijde (SvdH) method, and corresponding joint-level SvdH scores. This dataset enables consistent modeling of anatomical structures, lesion morphology, and clinical assessments. Experimental results demonstrate strong performance in full-hand bone segmentation, while pixel-level bone erosion segmentation remains challenging, underscoring the dataset’s critical value in advancing quantitative imaging analysis for RA.

Abstract

Rheumatoid arthritis (RA) assessment from hand radiographs requires multi-level analysis and modeling of anatomical structures and fine-grained local pathological changes. However, existing public resources do not support such unified multi-level analysis, often lacking full-hand coverage, fine-grained annotations, and consistent integration with clinical scoring systems. In particular, annotations that enable quantitative analysis of bone erosion (BE) remain scarce. RAM-H1200 contains 1,200 hand radiographs collected from six medical centers, with multi-level annotations including (i) whole-hand bone structure instance segmentation, (ii) pixel-level BE masks, (iii) SvdH-defined joint regions of interest, and (iv) joint-level SvdH scores for both BE and joint space narrowing (JSN). It is designed to evaluate whether models can jointly capture anatomical structure, localized erosive pathology, and clinically standardized RA severity from hand radiographs. The proposed BE masks enable, for the first time, quantitative BE analysis beyond coarse categorical grading by providing explicit spatial supervision for lesion extent and morphology. To our knowledge, RAM-H1200 is the first public large-scale benchmark that jointly supports whole-hand bone structure instance segmentation, pixel-level BE delineation, and clinically grounded joint-level SvdH scoring for both BE and JSN. Results across benchmark tasks show that anatomical modeling is substantially more mature than quantitative BE analysis: whole-hand bone segmentation achieves strong performance, whereas BE segmentation remains a major open challenge. By unifying anatomical structure modeling, quantitative lesion analysis, and clinically grounded SvdH scoring, RAM-H1200 provides a single benchmark for comprehensive RA analysis on hand radiographs.

Problem

• rheumatoid arthritis
• hand radiographs
• bone erosion
• multi-level annotation
• clinical scoring

Innovation

• bone erosion segmentation
• hand radiograph analysis
• SvdH scoring
• instance segmentation
• rheumatoid arthritis benchmark