Zheng Shuang, Ma Wenao, Mu Lin, He Kan, Cao Jianfeng, So Tiffany Y, Zhang Lei, Li Mingyang, Zhai Yanan, Liu Feng, Guo Shunlin, Yin Longlin, Zhao Liming, Wang Lei, Lee Heather H C, Jiang Wei, Niu Junqi, Gao Pujun, Dou Qi, Zhang Huimao
Department of Radiology, The First Hospital of Jilin University, Changchun, China.
Department of Computer Science and Engineering, The Chinese University of Hong Kong, Hong Kong, China.
iScience. 2025 Mar 17;28(4):112224. doi: 10.1016/j.isci.2025.112224. eCollection 2025 Apr 18.
Noninvasive methods for liver fibrosis staging are urgently needed due to its significance in predicting significant morbidity and mortality. In this study, we developed an automated DL-based segmentation and classification model (Model-C). Test-time adaptation was used to address data distribution shifts. We then established a deep learning-radiologist complementarity decision system (DRCDS) via a decision model determining whether to adopt Model-C's diagnosis or defer to radiologists. Model-C (AUCs of 0.89-0.92) outperformed models based on liver (AUCs: 0.84-0.90) or spleen (AUCs: 0.69-0.70). With test-time adaptation, the Obuchowski index values of Model-C in three external sets improved from 0.81, 0.73, and 0.73 to 0.85, 0.85, and 0.81. DRCDS performed slightly better than Model-C or senior radiologists, with 73.7%-92.0% of cases adopting Model-C's diagnosis. In conclusion, DRCDS could diagnose liver fibrosis with high accuracy. Additionally, we provided solutions to model generalization and human-machine complementarity issues in multi-classification problems.
由于肝纤维化分期在预测严重发病率和死亡率方面具有重要意义，因此迫切需要非侵入性方法。在本研究中，我们开发了一种基于深度学习的自动分割和分类模型（模型C）。测试时自适应被用于解决数据分布偏移问题。然后，我们通过一个决策模型建立了一个深度学习-放射科医生互补决策系统（DRCDS），该决策模型用于确定是采用模型C的诊断结果还是听从放射科医生的意见。模型C（曲线下面积为0.89 - 0.92）优于基于肝脏（曲线下面积：0.84 - 0.90）或脾脏（曲线下面积：0.69 - 0.70）的模型。通过测试时自适应，模型C在三个外部数据集上的奥布霍夫斯基指数值从0.81、0.73和0.73提高到了0.85、0.85和0.81。DRCDS的表现略优于模型C或资深放射科医生，73.7% - 92.0%的病例采用了模型C的诊断结果。总之，DRCDS能够高精度地诊断肝纤维化。此外，我们还为多分类问题中的模型泛化和人机互补问题提供了解决方案。
