Ria Francesco, Zhang Anru R, Lerebours Reginald, Erkanli Alaattin, Abadi Ehsan, Marin Daniele, Samei Ehsan
Carl E. Ravin Advanced Imaging Labs, Center for Virtual Imaging Trials, Department of Radiology, Duke University Health System, Durham, NC, USA.
Department of Biostatistics & Bioinformatics and Department of Computer Science, Duke University, Durham, NC, USA.
Commun Med (Lond). 2024 Dec 19;4(1):272. doi: 10.1038/s43856-024-00674-w.
Risk-versus-benefit optimization required a quantitative comparison of the two. The latter, directly related to effective diagnosis, can be associated to clinical risk. While many strategies have been developed to ascertain radiation risk, there has been a paucity of studies assessing clinical risk, thus limiting the optimization reach to achieve a minimum total risk to patients undergoing imaging examinations. In this study, we developed a mathematical framework for an imaging procedure total risk index considering both radiation and clinical risks based on specific tasks and investigated diseases.
The proposed model characterized total risk as the sum of radiation and clinical risks defined as functions of radiation burden, disease prevalence, false-positive rate, expected life-expectancy loss for misdiagnosis, and radiologist interpretative performance (i.e., AUC). The proposed total risk model was applied to a population of one million cases simulating a liver cancer scenario.
For all demographics, the clinical risk outweighs radiation risk by at least 400%. The optimization application indicates that optimizing typical abdominal CT exams should involve a radiation dose increase in over 90% of the cases, with the highest risk optimization potential in Asian population (24% total risk reduction; 306% increase) and lowest in Hispanic population (5% total risk reduction; 89% increase).
Framing risk-to-benefit assessment as a risk-versus-risk question, calculating both clinical and radiation risk using comparable units, allows a quantitative optimization of total risks in CT. The results highlight the dominance of clinical risk at typical CT examination dose levels, and that exaggerated dose reductions can even harm patients.
风险与获益的优化需要对两者进行定量比较。后者与有效诊断直接相关，可能与临床风险相关。虽然已经开发了许多策略来确定辐射风险，但评估临床风险的研究却很少，因此限制了优化范围，无法实现对接受影像检查患者的最低总风险。在本研究中，我们基于特定任务和所研究的疾病，开发了一个考虑辐射和临床风险的影像检查总风险指数的数学框架。
所提出的模型将总风险表征为辐射风险和临床风险之和，辐射风险和临床风险被定义为辐射负担、疾病患病率、假阳性率、误诊导致的预期寿命损失以及放射科医生解释性能（即曲线下面积）的函数。所提出的总风险模型应用于模拟肝癌情况的100万例病例群体。
对于所有人口统计学特征，临床风险比辐射风险至少高400%。优化应用表明，优化典型的腹部CT检查在超过90%的情况下应增加辐射剂量，亚洲人群的风险优化潜力最高（总风险降低24%；增加306%），西班牙裔人群最低（总风险降低5%；增加89%）。
将风险与获益评估构建为风险与风险的问题，使用可比单位计算临床风险和辐射风险，能够对CT中的总风险进行定量优化。结果突出了在典型CT检查剂量水平下临床风险的主导地位，并且过度降低剂量甚至可能对患者造成伤害。
