The U-Net was presented in 2015. With its straight-forward and successful architecture it quickly evolved to a commonly used benchmark in medical image segmentation. The adaptation of the U-Net to novel problems, however, comprises several degrees of freedom regarding the exact architecture, pre-processing, training and inference. These choices are not independent of each other and substantially impact the overall performance. The present paper introduces the nnU-Net ("nonew-Net"), which refers to a robust and self-adapting framework on the basis of 2D and 3D vanilla U-Nets. We argue the strong case for taking away superfluous bells and whistles of many proposed network designs and instead focus on the remaining aspects that make out the performance and generalizability of a method. We evaluate the nnU-Net in the context of the Medical Segmentation Decathlon challenge, which measures segmentation performance in ten disciplines comprising distinct entities, image modalities, image geometries and dataset sizes, with no manual adjustments between datasets allowed. At the time of manuscript submission, nnU-Net achieves the highest mean dice scores across all classes and seven phase 1 tasks (except class 1 in BrainTumour) in the online leaderboard of the challenge.

No abstract

For many applications, in particular in natural science, the task is to determine hidden system parameters from a set of measurements. Often, the forward process from parameter-to measurement-space is well-defined, whereas the inverse problem is ambiguous: multiple parameter sets can result in the same measurement. To fully characterize this ambiguity, the full posterior parameter distribution, conditioned on an observed measurement, has to be determined. We argue that a particular class of neural networks is well suited for this task -so-called Invertible Neural Networks (INNs). Unlike classical neural networks, which attempt to solve the ambiguous inverse problem directly, INNs focus on learning the forward process, using additional latent output variables to capture the information otherwise lost. Due to invertibility, a model of the corresponding inverse process is learned implicitly. Given a specific measurement and the distribution of the latent variables, the inverse pass of the INN provides the full posterior over parameter space. We prove theoretically and verify experimentally, on artificial data and real-world problems from medicine and astrophysics, that INNs are a powerful analysis tool to find multi-modalities in parameter space, uncover parameter correlations, and identify unrecoverable parameters.

PurposeMultispectral imaging can provide reflectance measurements at multiple spectral bands for each image pixel. These measurements can be used for estimation of important physiological parameters, such as oxygenation, which can provide indicators for the success of surgical treatment or the presence of abnormal tissue. The goal of this work was to develop a method to estimate physiological parameters in an accurate and rapid manner suited for modern high-resolution laparoscopic images.MethodsWhile previous methods for oxygenation estimation are based on either simple linear methods or complex model-based approaches exclusively suited for off-line processing, we propose a new approach that combines the high accuracy of model-based approaches with the speed and robustness of modern machine learning methods. Our concept is based on training random forest regressors using reflectance spectra generated with Monte Carlo simulations.ResultsAccording to extensive in silico and in vivo experiments, the method features higher accuracy and robustness than state-of-the-art online methods and is orders of magnitude faster than other nonlinear regression based methods.ConclusionOur current implementation allows for near real-time oxygenation estimation from megapixel multispectral images and is thus well suited for online tissue analysis.

Abstract. Multispectral imaging in laparoscopy can provide tissue reflectance measurements for each point in the image at multiple wavelengths of light. These reflectances encode information on important physiological parameters not visible to the naked eye. Fast decoding of the data during surgery, however, remains challenging. While modelbased methods suffer from inaccurate base assumptions, a major bottleneck related to competing machine learning-based solutions is the lack of labelled training data. In this paper, we address this issue with the first transfer learning-based method to physiological parameter estimation from multispectral images. It relies on a highly generic tissue model that aims to capture the full range of optical tissue parameters that can potentially be observed in vivo. Adaptation of the model to a specific clinical application based on unlabelled in vivo data is achieved using a new concept of domain adaptation that explicitly addresses the high variance often introduced by conventional covariance-shift correction methods. According to comprehensive in silico and in vivo experiments our approach enables accurate parameter estimation for various tissue types without the need for incorporating specific prior knowledge on optical properties and could thus pave the way for many exciting applications in multispectral laparoscopy.

Abstract. With the recent breakthrough success of machine learning based solutions for automatic image annotation, the availability of reference image annotations for algorithm training is one of the major bottlenecks in medical image segmentation and many other fields. Crowdsourcing has evolved as a valuable option for annotating large amounts of data while sparing the resources of experts, yet, segmentation of objects from scratch is relatively time-consuming and typically requires an initialization of the contour. The purpose of this paper is to investigate whether the concept of crowd-algorithm collaboration can be used to simultaneously (1) speed up crowd annotation and (2) improve algorithm performance based on the feedback of the crowd. Our contribution in this context is two-fold: Using benchmarking data from the MICCAI 2015 endoscopic vision challenge we show that atlas forests extended by a novel superpixel-based confidence measure are well-suited for medical instrument segmentation in laparoscopic video data. We further demonstrate that the new algorithm and the crowd can mutually benefit from each other in a collaborative annotation process. Our method can be adapted to various applications and thus holds high potential to be used for large-scale low-cost data annotation.

This paper significantly enhances the state of art in automatic labeling of endoscopic videos by introducing the use of the confidence metric, and by being the first study to use MI data for in vivo laparoscopic tissue classification. The data of our experiments will be released as the first in vivo MI dataset upon publication of this paper.

Purpose: Optical imaging is evolving as a key technique for advanced sensing in the operating room. Recent research has shown that machine learning algorithms can be used to address the inverse problem of converting pixel-wise multispectral reflectance measurements to underlying tissue parameters, such as oxygenation. Assessment of the specific hardware used in conjunction with such algorithms, however, has not properly addressed the possibility that the problem may be illposed. Methods: We present a novel approach to the assessment of optical imaging modalities, which is sensitive to the different types of uncertainties that may occur when inferring tissue parameters. Based on the concept of invertible neural networks, our framework goes beyond point estimates and maps each multispectral measurement to a full posterior probability distribution which is capable of representing ambiguity in the solution via multiple modes. Performance metrics for a hardware setup can then be computed from the characteristics of the posteriors. Results: Application of the assessment framework to the specific use case of camera selection for physiological parameter estimation yields the following insights: (1) Estimation of tissue oxygenation from multispectral images is a well-posed problem, while (2) blood volume fraction may not be recovered without ambiguity. (3) In general, ambiguity may be reduced by increasing the number of spectral bands in the camera. Conclusion: Our method could help to optimize optical camera design in an application-specific manner.