DiffVP: Differential Visual Semantic Prompting for LLM-Based CT Report Generation

TL;DR

DiffVP introduces a novel approach for CT report generation by focusing on high-level semantic differences between scans, improving accuracy and clinical relevance without explicit lesion localization.

Contribution

The paper proposes Differential Visual Prompting (DiffVP), a new method that leverages semantic scan-to-reference differences to enhance LLM-based CT report generation.

Findings

Outperforms prior methods on two large-scale benchmarks.

Improves BLEU scores significantly (+10.98 and +4.36).

Boosts clinical efficacy with an F1 score of 0.421.

Abstract

While large language models (LLMs) have advanced CT report generation, existing methods typically encode 3D volumes holistically, failing to distinguish informative cues from redundant anatomical background. Inspired by radiological cognitive subtraction, we propose Differential Visual Prompting (DiffVP), which conditions report generation on explicit, high-level semantic scan-to-reference differences rather than solely on absolute visual features. DiffVP employs a hierarchical difference extractor to capture complementary global and local semantic discrepancies into a shared latent space, along with a difference-to-prompt generator that transforms these signals into learnable visual prefix tokens for LLM conditioning. These difference prompts serve as structured conditioning signals that implicitly suppress invariant anatomy while amplifying diagnostically relevant visual evidence, thereby facilitating accurate report generation without explicit lesion localization. On two large-scale benchmarks, DiffVP consistently outperforms prior methods, improving the average BLEU-1-4 by +10.98 and +4.36, respectively, and further boosts clinical efficacy on RadGenome-ChestCT (F1 score 0.421). All codes will be released at https://github.com/ArielTYH/DiffVP/.