基于感知的非均匀彩色映射在CT阅片中的应用评估
本研究开发了一种支持交互式探索的感知彩色映射软件框架,并通过对比剂稀释序列和肝脏体模CT扫描进行验证。结果显示非均匀彩色映射能稳定呈现视觉表现,且在主观视觉分析中显著优于标准均匀显示方法。
要点
- 非均匀彩色映射可减少手动窗宽窗位调整,可能提高CT阅片效率。
- 该方法在体模实验中视觉评分显著优于传统灰阶与均匀彩色显示。
- 彩色映射的稳定性在不同剂量和重建设置下得到验证。
核心结论
- 非均匀彩色映射比均匀彩色映射更符合人类感知,在体模视觉评估中优于标准均匀显示方法。
- 该方法可能有助于CT图像中组织与结构的视觉区分。
研究设计与样本
- 研究类型为方法开发与体模验证研究。
- 开发了支持交互式非均匀彩色显示的专用研究用DICOM查看器。
- 验证包括对比剂稀释序列与临床查看器对比,以及不同剂量和重建设置下的肝脏体模CT扫描。
- 肝脏体模中包含预定义病灶。
关键影像表现/征象
- 对比剂稀释序列定量验证确认了CT值到颜色的正确映射。
- 非均匀彩色映射在肝脏体模CT扫描中产生了稳定的视觉表现。
- 在基于体模的读片评估中,非均匀彩色映射在主观视觉分析中被显著偏好。
临床应用与阅片要点
- 非均匀彩色映射是一种有前景的CT显示替代方法,值得在更多临床场景中验证。
- 该方法可能减少对频繁手动调整灰阶窗宽窗位的依赖。
局限性
- 本研究仅基于体模实验,未在真实临床病例中验证。
- 主观视觉评估存在固有偏倚风险。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
BACKGROUND: With modern CT technology offering faster acquisition, lower radiation exposure, higher resolution, and advanced detector designs, routine image interpretation still relies largely on gray‑scale windowing, which by design requires frequent manual adjustment. Existing color mappings rely on uniform mathematical gradients and offer no clear advantage over gray‑scale display for tissue or structure perception.
OBJECTIVE: To develop and evaluate a software framework supporting interactive exploration of perception‑based color mapping.
MATERIALS AND METHODS: A custom research DICOM viewer enabling interactive nonuniform color display was developed. Verification included a contrast‑agent dilution series compared with a clinical viewer and liver phantom CT scans with predefined lesions acquired at varying dose levels and reconstruction settings. Differentiation of tissues and structures in the liver phantom was examined by visual rating after application of different color mappings.
RESULTS: Quantitative verification using a contrast-agent dilution series confirmed correct mapping of CT values to color. In a phantom-based reader evaluation using liver phantom CT scans, nonuniform color mapping produced stable visual representations and was significantly preferred in subjective visual analysis.
CONCLUSIONS: Unlike uniform color mappings, the proposed approach uses nonuniform color ranges to better match human perception. In CT, this may facilitate visual separation of tissues and structures and was preferred in phantom‑based visual assessment compared with standard uniform display methods.
引用
- Landolt, Euler, Froehlich, et al. Perception‑based, Nonuniform Color Mapping for CT Reading. Invest Radiol. 2026. PMID: 42548170. doi:10.1097/RLI.0000000000001306
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。