arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2005.04454cs.LGeess.IVstat.ML
收藏

无监督体育训练的特定领域损失设计:一种建模医疗ML解决方案的新方法

Domain-specific loss design for unsupervised physical training: A new approach to modeling medical ML solutions

Technische Universität München(慕尼黑工业大学);Carl Zeiss Meditec AG(卡尔蔡司医疗技术公司);Vienna Hanusch Hospital(维也纳哈努施医院);Johns Hopkins University(约翰斯·霍普金斯大学);Ludwig-Maximilians Universität München(慕尼黑大学)

Hendrik Burwinkel, Holger Matz, Stefan Saur, Christoph Hauger, Ayse Mine Evren, Nino Hirnschall, Oliver Findl, Nassir Navab, Seyed-Ahmad Ahmadi

摘要:

今天,白内障手术是世界上最常进行的眼科手术。白内障是一种正在发展的人类晶状体混浊,是世界上最常见的致盲原因。在手术期间,晶状体被移除并被人工眼内晶状体(IOL)替换。为了防止患者在手术后需要强大的视觉辅助,植入IOL的光学特性的精确预测至关重要。已经有很多活动致力于开发从OCT设备获得的生物眼睛数据预测这些特性的方法,最近也通过采用机器学习。他们要么只考虑生物特征数据,要么只考虑物理模型,但很少同时考虑两者,并且经常忽略IOL的几何形状。在这项工作中,我们提出了OpticNet,一种新的光学折射网络,损失函数,和训练计划,这是无监督的,特定于领域的,和物理动机。我们推导出一个精确的光传播眼睛模型,使用单射线光线跟踪和制定一个可微的损失函数,反向传播到网络的物理梯度。此外,我们提出了一种新的迁移学习程序,它允许对物理模型进行无监督训练,并对一组真实IOL患者病例进行网络微调。我们表明,我们的网络不仅优于用标准程序训练的系统,而且当在两个生物特征数据集上进行比较时,我们的方法在IOL计算中的表现优于目前最先进的技术。

英文摘要:

Today, cataract surgery is the most frequently performed ophthalmic surgery in the world. The cataract, a developing opacity of the human eye lens, constitutes the world's most frequent cause for blindness. During surgery, the lens is removed and replaced by an artificial intraocular lens (IOL). To prevent patients from needing strong visual aids after surgery, a precise prediction of the optical properties of the inserted IOL is crucial. There has been lots of activity towards developing methods to predict these properties from biometric eye data obtained by OCT devices, recently also by employing machine learning. They consider either only biometric data or physical models, but rarely both, and often neglect the IOL geometry. In this work, we propose OpticNet, a novel optical refraction network, loss function, and training scheme which is unsupervised, domain-specific, and physically motivated. We derive a precise light propagation eye model using single-ray raytracing and formulate a differentiable loss function that back-propagates physical gradients into the network. Further, we propose a new transfer learning procedure, which allows unsupervised training on the physical model and fine-tuning of the network on a cohort of real IOL patient cases. We show that our network is not only superior to systems trained with standard procedures but also that our method outperforms the current state of the art in IOL calculation when compared on two biometric data sets.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑