(The talk will be given in English)

Speaker:     Dr. Shai Vardi
                   Caltech, USA

Monday, December 18^th, 2017
15:00 - 16:00

Room 011, Kitot Bldg., Faculty of Engineering

Meeting the challenges of massive networks and systems

Abstract

Massive systems and networks have become ubiquitous. While there is a remarkable amount of work on analyzing and designing algorithms for smaller systems, the vast majority of it simply does not scale: an algorithm that takes seconds to run on a system with thousands of nodes might take weeks on a system with billions. New ideas are required if we hope to have the same success with massive systems as we do with smaller ones. The field of Local computation algorithms (LCAs), which I founded together with Ronitt Rubinfeld, Gil Tamir and Ning Xie, provides a rigorous framework for solving problems on huge systems. I will present one technique for designing LCAs: a reduction to online algorithms, and show how it can be used to design fast and robust distributed solvers for linear and convex programs. I will also describe how we can use insights from this technique to design approximate solvers for linear programs that have hitherto been unsolvable due to their size.

Bio
Shai Vardi is a Linde Postdoctoral Fellow at the Social and Information Sciences Laboratory at the California Institute of Technology, hosted by Adam Wierman. He spent the Fall semester of 2016-2017 at the Simons Institute for the Theory of Computation, and was previously a postdoc in the Weizmann Institute of Science, advised by Uri Feige. He received his MSc and PhD in Computer Science from Tel Aviv University, advised by Ronitt Rubinfeld and Yishay Mansour respectively. Shai received the Google European Fellowship for Game Theory for 2012-2015, and the iCORE Algorithms Postdoctoral Scholarship in 2015. He also has a diploma in jazz performance from the Rimon School of Jazz and Contemporary Music

מזמינים אתכם להדלקת נר ראשון ומסיבת חנוכה בחסות חברת Altair 

11:00-13:00 - נפגשים בלובי ביניין כיתות 011 לשמוע ולהכיר מקרוב את מהנדסי Altair

13:00-14:00 - איגור טובברג, מנהל מוצר, ידבר על LTE, LOT וחדקונית הדקל.

14:00 - הדלקת נר ראשון של חנוכה

הגיעו עם קו"ח מעודכנים!

Speaker: Itzik Avital

M.Sc. student under the supervision of Prof. Nahum Kiryati and Dr. Arnaldo Mayer

Wednesday, December 27^th 2017 at 15:30

Room 011, Kitot Bldg., Faculty of Engineering

3D Multi-Modal FCNN for Automatic Segmentation of Anatomical ROIs for Seeding Pre-surgical Brain Tractography

Abstract

White matter tractography is a 3D modeling of neural tracts in the brain. It is an important tool for neuro-surgical planning and navigation. Stringent pre-operative time-constraints and limited availability of neuroanatomy experts indicate that pre-surgical tractography would clearly benefit from automation tools. Also, multi-modal methods, which use a number of different type inputs, have recently been found to be beneficial for many computer vision tasks. In this work, we propose a deep learning based method for automatic segmentation of seeding region-of-interest (SROI) for tractography. Moreover, we show that using the information from the white matter orientation patterns (PDD, Principal Diffusion Direction) achieves better results than using only the anatomical information (MRI scan). Also, we propose a novel approach for exploiting information from a number of inputs. This approach can be implemented by any encoder-decoder architecture. In this work, the motor, optic and arcuate tracts (right and left) are considered.

   We defined five 3D encoder-decoder FCNNs (Fully Convolutional Neural Network). The first is fed with an MRI scan. The second is fed with a PDD map. The third is fed with an MRI scan and a PDD map which are concatenated as one input volume with 4 channels (early fusion). The fourth is composed of two streams, which are fused right before the two last layers; one is fed with an MRI scan and the other with a PDD map (late fusion). The fifth is an implementation of our novel approach. It is composed of two encoders (one for each modality) and one shared decoder (intermediate fusion). Each architecture has been trained for each tract type (arcuate, motor and visual) and over 5 cross-validation folds. A database consisting 75 cases, referred to Sheba Medical Center for brain tumor removal surgery, was used for the experiments. The presence of a brain tumor is clearly apparent in most of them which makes the task much more complex.

    The superiority of the proposed architecture and the necessity of the PDD information for SROIs segmentation are confirmed by a number of metrics. Fully automatic tractography has been obtained using the automatic segmented SROIs and has been successfully validated against manual tractography. The proposed method provides a promising approach for automatic SROIs segmentation for tractography and can be used as an automatic tool.