Layer Separation: Adjustable Joint Space Width Images Synthesis in Conventional Radiography
This work addresses data imbalance, limited variety, and annotation difficulties in computer-aided diagnostic systems for rheumatoid arthritis, which is incremental as it builds on existing synthesis methods for a specific medical imaging domain.
The paper tackled the challenge of data quality issues in deep learning-based radiological CAD systems for joint space width analysis in rheumatoid arthritis by proposing Layer Separation Networks to synthesize adjustable JSW images, resulting in synthetic images that closely resemble real radiographs and significantly enhance downstream task performance.
Rheumatoid arthritis (RA) is a chronic autoimmune disease characterized by joint inflammation and progressive structural damage. Joint space width (JSW) is a critical indicator in conventional radiography for evaluating disease progression, which has become a prominent research topic in computer-aided diagnostic (CAD) systems. However, deep learning-based radiological CAD systems for JSW analysis face significant challenges in data quality, including data imbalance, limited variety, and annotation difficulties. This work introduced a challenging image synthesis scenario and proposed Layer Separation Networks (LSN) to accurately separate the soft tissue layer, the upper bone layer, and the lower bone layer in conventional radiographs of finger joints. Using these layers, the adjustable JSW images can be synthesized to address data quality challenges and achieve ground truth (GT) generation. Experimental results demonstrated that LSN-based synthetic images closely resemble real radiographs, and significantly enhanced the performance in downstream tasks. The code and dataset will be available.