No abstract

The level set based deformable models (LDM) are commonly used for medical image segmentation. However, they rely on a handcrafted curve evolution velocity that needs to be adapted for each segmentation task. The Convolutional Neural Networks (CNN) address this issue by learning robust features in a supervised end-to-end manner. However, CNNs employ millions of network parameters which require a large amount of training data to prevent over-fitting and also increases its memory requirement and computation time during testing. Moreover, since CNNs pose segmentation as a region-based pixel labeling, they cannot explicitly model the high-level dependencies between the points on the object boundary to preserve its overall shape, smoothness or the regional homogeneity within and outside the boundary. We present a Recurrent Neural Network (RNN) based solution called the RACE-net to address the above issues. RACE-net models a generalized LDM evolving under a constant and mean curvature velocity. At each time-step, the curve evolution velocities are approximated using a feed-forward architecture inspired by the multi-scale image pyramid. RACE-net allows the curve evolution velocities to be learned in an end-to-end manner while minimizing the number of network parameters, computation time and memory requirements. The RACE-net was validated on three different segmentation tasks: Optic disc and cup in color fundus images, cell nuclei in histopathological images and the left atrium in cardiac MRI volumes. Assessment on public datasets was seen to yield high Dice values between 0.87 and 0.97 which illustrates its utility as a generic, off-the-shelf architecture for biomedical segmentation.

No abstract

An automated system was presented for glaucoma detection from color fundus photographs. The overall evaluation results indicated that the presented system was comparable in performance to glaucoma classification by a manual grader solely based on fundus image examination.

The explosive growth of data and its related energy consumption is pushing the need to develop energyefficient brain-inspired schemes and materials for data processing and storage. Here, we demonstrate experimentally that Co/Pt films can be used as artificial synapses by manipulating their magnetization state using circularly-polarized ultrashort optical pulses at room temperature. We also show an efficient implementation of supervised perceptron learning on an opto-magnetic neural network, built from such magnetic synapses. Importantly, we demonstrate that the optimization of synaptic weights can be achieved using a global feedback mechanism, such that the learning does not rely on external storage or additional optimization schemes. These results suggest there is high potential for realizing artificial neural networks using optically-controlled magnetization in technologically relevant materials, that can learn not only fast but also energy-efficient.The rapid growth of modern information and communication technology (ICT) has led to an enormous increase in energy consumption, which is already now around 7% of the global electrical energy production. Owing to the inherently energy-efficient brain-inspired computing principles, implementing such neuromorphic architectures offers enormous potential to dramatically reduce the energy consumption of ICT. Magnetic materials are already at the center of computing today, due to their ability to store information within the direction of magnetic moments in a non-volatile and rewritable way. In recent years, tremendous progress has been made in controlling magnetism with femtosecond optical pulses, including demonstrations of record-breaking fast write-read events 1 , operation in technologically relevant materials such as Co/Pt 2 and enabling magnetic recording that is not only much faster but also exhibits a projected heat load of only 22 aJ per magnetic bit 3 . These demonstrations suggest that all-optical manipulation of magnetism offers great potential to realize neural networks that can be trained not only much faster but also much more efficiently than all-electrical material implementations under development today 4-9 . In particular, similar as for purely optical neural networks 10-12 operation proceeds ultrafast and at very low energy cost. Moreover, combining this with magnetism integrates inherent non-volatility and optical adaptability, which potentially yields fast and energy-efficient learning as well. However, so far it has not been demonstrated that it is possible to exploit optical control of magnetism for brain-inspired computing. Here, we study the control of the magnetization state of Co/Pt thin films in response to picosecond optical pulses. We experimentally demonstrate that the cumulative all-optical switching process in these materials 13,14 provides an energy-efficient mechanism for realizing artificial synapses, in which the internal state of the synapse is stored with non-volatility and can be controlled continuously and reversibly using the h...

No abstract

Abstract. We present a novel framework for depth based optic cup boundary extraction from a single 2D color fundus photograph per eye. Multiple depth estimates from shading, color and texture gradients in the image are correlated with Optical Coherence Tomography (OCT) based depth using a coupled sparse dictionary, trained on image-depth pairs. Finally, a Markov Random Field is formulated on the depth map to model the relative depth and discontinuity at the cup boundary. Leaveone-out validation of depth estimation on the INSPIRE dataset gave average correlation coefficient of 0.80. Our cup segmentation outperforms several state-of-the-art methods on the DRISHTI-GS dataset with an average F-score of 0.81 and boundary-error of 21.21 pixels on test set against manual expert markings. Evaluation on an additional set of 28 images against OCT scanner provided groundtruth showed an average rms error of 0.11 on Cup-Disk diameter and 0.19 on Cup-disk area ratios.

Population screening for sight threatening diseases based on fundus imaging is in place or being considered worldwide. Most existing programs are focussed on a specific disease and are based on manual reading of images, though automated image analysis based solutions are being developed. Exudates