Ex vivo lung perfusion (EVLP) enables advanced assessment of human lungs for transplant suitability. We developed a convolutional neural network (CNN)-based approach to analyze the largest cohort of isolated lung radiographs to date. CNNs were trained to process 1300 longitudinal radiographs from n = 650 clinical EVLP cases. Latent features were transformed into principal components (PC) and correlated with known radiographic findings. PCs were combined with physiological data to classify clinical outcomes: (1) recipient time to extubation of <72 h, (2) ≥ 72 h, and (3) lungs unsuitable for transplantation. The top PC was significantly correlated with infiltration (Spearman R: 0·72, p < 0·0001), and adding radiographic PCs significantly improved the discrimination for clinical outcomes (Accuracy: 73 vs 78%, p = 0·014). CNN-derived radiographic lung features therefore add substantial value to the current assessments. This approach can be adopted by EVLP centers worldwide to harness radiographic information without requiring real-time radiological expertise.

离体肺灌注 (EVLP) 可以对人肺的移植适宜性进行高级评估。我们开发了一种基于卷积神经网络 (CNN) 的方法来分析迄今为止最大的一组孤立肺 X 光照片。 CNN 经过训练，可以处理来自n = 650 例临床 EVLP 病例的 1300 张纵向放射线照片。潜在特征被转化为主成分（PC）并与已知的射线照相结果相关联。 PC 与生理数据相结合，对临床结果进行分类：(1) 受者拔管时间为 <72 小时，(2) ≥ 72 小时，(3) 肺不适合移植。顶部 PC 与浸润显着相关（Spearman R：0·72， p < 0·0001），并且添加放射学 PC 显着改善了对临床结果的辨别（准确度：73 vs 78%， p = 0·014）。因此，CNN 衍生的肺部放射学特征为当前的评估增添了巨大的价值。世界各地的 EVLP 中心都可以采用这种方法来利用放射学信息，而无需实时放射学专业知识。