Distributed deep learning for robust multi-site segmentation of CT imaging after traumatic brain injury.
Distributed deep learning for robust multi-site segmentation of CT imaging after traumatic brain injury.
复制标题
分布式深度学习，用于脑外伤后 CT 成像的稳健多部位分割。
DOI：
10.1117/12.2511997
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
Pham,DzungL
中科院分区：
文献类型：
--
作者：
Remedios,Samuel;Roy,Snehashis;Blaber,Justin;Bermudez,Camilo;Nath,Vishwesh;Patel,MayurB;Butman,JohnA;Landman,BennettA;Pham,DzungL
机器学习模型在医学成像领域变得司空见惯，随着这些方法的出现，对更多数据的需求也在不断增加。然而，为了保护患者的匿名性，在机构之间传输受保护的健康信息(PHI)通常是不切实际的或被禁止的。此外，由于某些研究的性质，可能没有可用于训练模型的大型公共数据集。为了解决这个难题，我们分析了在不同站点之间传输模型本身而不是数据的有效性。通过这样做，我们实现了两个目标：1)模型获得了对通常无法获得的更大数据集的训练；2)模型对来自不同位置的数据进行了训练，从而更好地泛化。在这篇文章中，我们使用来自美国国立卫生研究院(NIH)和范德比尔特大学医学中心(VUMC)的不同数据集实施多站点学习，而不影响PHI。三个神经网络被训练来收敛于计算机断层扫描(CT)脑血肿分割任务：一个仅使用NIH数据，一个仅使用VUMC数据，以及一个在NIH和VUMC数据之间交替的多站点模型。使用多部位模型得到的病变掩膜的平均Dice相似系数为0.64，自动分割的血肿体积与手动分割的血肿体积相关，皮尔逊相关系数为0.87，分别比单部位模型对应的模型提高了8%和5%。
Machine learning models are becoming commonplace in the domain of medical imaging, and with these methods comes an ever-increasing need for more data. However, to preserve patient anonymity it is frequently impractical or prohibited to transfer protected health information (PHI) between institutions. Additionally, due to the nature of some studies, there may not be a large public dataset available on which to train models. To address this conundrum, we analyze the efficacy of transferring the model itself in lieu of data between different sites. By doing so we accomplish two goals: 1) the model gains access to training on a larger dataset that it could not normally obtain and 2) the model better generalizes, having trained on data from separate locations. In this paper, we implement multi-site learning with disparate datasets from the National Institutes of Health (NIH) and Vanderbilt University Medical Center (VUMC) without compromising PHI. Three neural networks are trained to convergence on a computed tomography (CT) brain hematoma segmentation task: one only with NIH data, one only with VUMC data, and one multi-site model alternating between NIH and VUMC data. Resultant lesion masks with the multi-site model attain an average Dice similarity coefficient of 0.64 and the automatically segmented hematoma volumes correlate to those done manually with a Pearson correlation coefficient of 0.87, corresponding to an 8% and 5% improvement, respectively, over the single-site model counterparts.