On the 11th of June, 2009, the science Faculty day held a competition between MSc and PhD students who presented posters about their research. Almost 150 posters were presented from all faculty departments; Thirteen MSc and PhD students of the CSE school presented their work. The best four posters among the CSE posters were chosen, and awarded for their work. Here they are:

• Gaddi Blumrosen, Bio-engineering dept.
• Elad Eban, CS dept.
• Moti Freiman, CS dept.
• Yael Pritch, CS dept.

New wearable body sensor for continuous diagnosis of internal tissue bleeding (Gaddi Blumrosen)

Continuous monitoring of internal bleeding in tissue is essential for assessment and treatment of medical conditions. In this study we present a new body sensor for diagnosis of internal tissue bleeding based on contact multi-frequency electromagnetic inductive measurements of the tissue. This sensor can be used in emergency medicine in the field, in a hospital, or during daily life for patients at risk of bleeding. Unlike other sensors that detect bleeding, this sensor is portable, low cost, non-invasive and does not require galvanic coupling between the electrode and the tissue under measurement. Our study includes description of the new body sensor, criteria for data processing, basic diagnostic algorithms and calibration techniques. We show feasibility of this new sensor with a sensor prototype and diagnosis of an experimental simulation of internal bleeding and hypoperfusion conditions in the brain. This work is part of a joint efford of two groups: in Hebrew University (and Berekly), led by Professor Boris Rubinsky and in Mexico, led By Proffesor Cesar Gonzelles. These days, we continue the work in animals, and hopefully will be able to start expiremnts with people in future.

We simulated head injury bleeding for first order proof of the feasibility of the system.We used a sensor prototype and a head model. We supplied a cosine signal of approximately 10mA at five frequencies of 10, 40, 110, 160 MHz and 1GHzFluid volumes of physiological saline in 20ml decrement and increments were injected or removed to simulate different hypoperfusion or hyperperfusion levels respectively.

The poster

Interactive Proofs for Quantum Computations (Elad Eban)

The widely held belief that BQP (The class of problems solvable in polynomial time by a quantum computer) strictly contains BPP (the class of problems solvable in polynomial time by a probabilistic TM) raises fundamental questions: Upcoming generations of quantum computers might already be too large to be simulated classically. Is it possible to experimentally test that these systems perform as they should, if we cannot efficiently compute predictions for their behavior? If computing predictions for Quantum Mechanics requires exponential resources, is Quantum Mechanics a falsifiable theory? In cryptographic settings, an untrusted future company wants to sell a quantum computer or perform a delegated quantum computation. Can the customer be convinced of correctness without the ability to compare results to predictions? To provide positive answers to these questions, we define Quantum Prover Interactive Proofs (QPIP). Whereas in standard Interactive Proofs [GMR85] the prover is computationally unbounded, here our prover is in BQP, representing a quantum computer. The verifier models our current computational capabilities: it is a BPP machine, with access to few qubits.
Interactive Proofs For Quantum Computations

A blind protocols is one where the prover gains no information about the function being computed or the input.

The poster

Classification of suspected liver metastases using fMRI images: a machine learning approach (Moti Freiman)

We present a machine-learning approach to the interactive classification of suspected liver metastases in fMRI images. The method uses fMRI-based statistical modeling to characterize colorectal hepatic metastases and follow their early hemodynamical changes. Changes in hepatic hemodynamics are evaluated from T2 -W fMRI images acquired during the breathing of air, air-CO2, and carbogen. A classification model is build to differentiate between tumors and healthy liver tissues. To validate our method, a model was built from 29 mice datasets, and used to classify suspicious regions in 16 new datasets of healthy subjects or subjects with metastases in earlier growth phases. Our experimental results on mice yielded an accuracy of 78% with high precision (88%). This suggests that the method can provide a useful aid for early detection of liver metastases.
Project webpage

Example of a partitioned liver from the early growth phase. The suspected region is encircled by the yellow circle. The images in upper line are the anatomical image and functional maps from the early stage. The images in the middle are the partitioned (left) and classified (right) image colored according to the results. The image in the bottom is the anatomical image from the advanced growth phase. The metastasis region corresponds to the suspected region is encircled with a red circle.

The poster

Clustered Synopsis of Surveillance Video (Yael Pritch)

With: Sarit Ratovitch, Avishai Hendel and Shmuel Peleg.

Yael Pritch	Avishai Hendel	Sarit Ratovitch

Millions of surveillance cameras record video around the clock, producing huge video archives. Even when a video archive is known to include critical activities, finding them is like finding a needle in a haystack, making the archive almost worthless. Two main approaches were proposed to address this problem: action recognition and video summarization. Methods for automatic detection of activities still face problems in many scenarios. The video synopsis approach to video summarization is very effective, but may produce confusing summaries by the simultaneous display of multiple activities. A new methodology for the generation of short and coherent video summaries is presented, based on clustering of similar activities. Objects with similar activities are easy to watch simultaneously, and outliers can be spotted instantly. Clustered synopsis is also suitable for efficient creation of ground truth data.
Clustered Synopsis of Surveillance Video (project page). (With: Sarit Ratovitch Avishai Hendel and Shmuel Peleg ).