We present a novel 3-D warping technique for the estimation of 4-D seismic time-shift. This unsupervised method provides a diffeomorphic 3-D time shift field that includes uncertainties, therefore, it does not need prior time-shift data to be trained. This results in a widely applicable method in time-lapse seismic data analysis that is not implicitly biased by supervised time-shifts from other methods. We explore the generalization of the method to unseen data both in the same geological setting and in a different field, where the generalization error stays constant and within an acceptable range across test cases. We further explore upsampling of the warp field from a smaller network to decrease computational cost and see some deterioration of the warp field quality as a result. This method provides an accurate 3-D seismic registration method, where the heavy computation can be preexecuted and the inference of the network taking seconds on consumer hardware.

OBJECTIVE: In this study, wepropose an automatic diagnostic algorithm for detecting otitis media based on wideband tympanometry measurements. METHODS: We develop a convolutional neural network for classification of otitis media based on the analysis of the wideband tympanogram. Saliency maps are computed to gain insight into the decision process of the convolutional neural network. Finally, we attempt to distinguish between otitis media with effusion and acute otitis media, a clinical subclassification important for the choice of treatment. RESULTS: The approach shows high performance on the overall otitis media detection with an accuracy of 92.6%. However, the approach is not able to distinguish between specific types of otitis media. CONCLUSION: Out approach can detect otitis media with high accuracy and the wideband tympanogram holds more diagnostic information than the commonly used techniques wideband absorbance measurements and simple tympanograms. SIGNIFICANCE: This study shows how advanced deep learning methods enable automatic diagnosis of otitis media based on wideband tympanometry measurements, which could become a valuable diagnostic tool.

We introduce a parallel version of the Quadratic Pseudo-Boolean Optimization (QPBO) algorithm for solving binary optimization tasks, such as image segmentation. The original QPBO implementation by Kolmogorov and Rother relies on the Boykov-Kolmogorov (BK) maxflow/mincut algorithm and performs well for many image analysis tasks. However, the serial nature of their QPBO algorithm results in poor utilization of modern hardware. By redesigning the QPBO algorithm to work with parallel maxflow/mincut algorithms, we significantly reduce solve time of large optimization tasks. We compare our parallel QPBO implementation to other state-of-the-art solvers and benchmark them on two large segmentation tasks and a substantial set of small segmentation tasks. The results show that our parallel QPBO algorithm is over 20 times faster than the serial QPBO algorithm on the large tasks and over three times faster for the majority of the small tasks. Although we focus on image segmentation, our algorithm is generic and can be used for any QPBO problem. Our implementation and experimental results are available at DOI: 10.5281/zenodo.5201620

In this study, we propose an automatic diagnostic algorithm for detecting otitis media based on otoscopy images of the tympanic membrane. A total of 1336 images were assessed by a medical specialist into three diagnostic groups: acute otitis media, otitis media with effusion, and no effusion. To provide proper treatment and care and limit the use of unnecessary antibiotics, it is crucial to correctly detect tympanic membrane abnormalities, and to distinguish between acute otitis media and otitis media with effusion. The proposed approach for this classification task is based on deep metric learning, and this study compares the performance of different distance-based metric loss functions. Contrastive loss, triplet loss and multi-class N-pair loss are employed, and compared with the performance of standard cross-entropy and class-weighted cross-entropy classification networks. Triplet loss achieves high precision on a highly imbalanced data set, and the deep metric methods provide useful insight into the decision making of a neural network. The results are comparable to the best clinical experts and paves the way for more accurate and operator-independent diagnosis of otitis media.

We introduce the novel concept of a Sparse Layered Graph (SLG) for s-t graph cut segmentation of image data. The concept is based on the widely used Ishikawa layered technique for multi-object segmentation, which allows explicit object interactions, such as containment and exclusion with margins. However, the spatial complexity of the Ishikawa technique limits its use for many segmentation problems. To solve this issue, we formulate a general method for adding containment and exclusion interaction constraints to layered graphs. Given some prior knowledge, we can create a SLG, which is often orders of magnitude smaller than traditional Ishikawa graphs, with identical segmentation results. This allows us to solve many problems that could previously not be solved using general graph cut algorithms. We then propose three algorithms for further reducing the spatial complexity of SLGs, by using ordered multi-column graphs. In our experiments, we show that SLGs, and in particular ordered multi-column SLGs, can produce high-quality segmentation results using extremely simple data terms. We also show the scalability of ordered multi-column SLGs, by segmenting a high-resolution volume with several hundred interacting objects.