Hybrid Retinal Prosthesis: High-Resolution Electrode Array Integrated with Neurons for Restoration of Sight
Yossi Mandel, MD, PhD
Ophthalmic Science and Engineering Lab
Faculty of Life Sciences, School of Optometry and Visual Science and Institute for Nanotechnology and Advanced Materials (BINA), Bar-Ilan University, Ramat-Gan, Israel. 

Bio
Dr. Yossi Mandel is an assistant Professor in the School of Optometry and Visual Science and Institute for Nanotechnology and Advanced Materials (BINA), Bar-Ilan University, Israel. He is a medical doctor and ophthalmic surgeon and holds a PhD in bioengineering from the Hebrew University, Jerusalem. He did his post-doc with Daniel Palanker at Stanford University and he is now heading the Ophthalmic Science and Engineering Lab at Bar-Ilan University. His research is focused on various ophthalmic technologies aimed at restoration of sight, such as retinal prosthesis, stem cell and combination of these modalities.

Abstract
Vision restoration in patients with outer retinal degenerative diseases, such as Age-related Macular Degeneration and Retinitis Pigmentosa can be achieved by bypassing the degenerated photoreceptors and the electrical stimulation of the relatively well-preserved inner retina through electrode implants. Although current retinal prostheses have been shown to provide useful vision in blind patients, the obtained visual acuity and quality are still relatively low. Network mediated, rather than direct ganglion cell, activation can potentially facilitate the utilization of the complex computational processes performed by the inner retinal circuits. However, inherent limitations of current retinal prosthetic technologies, make it very challenging to mimic natural vision for better restoration of sight.
Firstly, increasing the electrode density for achieving high visual acuity is limited by the distance between the electrodes and the target neurons, which is currently a few tens to hundreds of microns. Secondly, the high electrode-neural distance results in relatively large activation charge thresholds, which raises the need for a pulsed as opposed to continuous current injection. Thus, the bipolar cells are operated in a pulsed rather than a graded potential fashion which provides the natural visual system with its unrivalled dynamic range and sensitivity. More importantly, the direct electrical stimulation of the bipolar cells fails to preserve selectivity of the specific retinal circuitry (such as ON and OFF pathways), resulting in distorted information being delivered to the brain.

We propose a paradigm shift toward sight restoration with a hybrid retinal prosthesis aimed at overcoming the aforementioned limitations by better mimicking natural vision. The hybrid implant is composed of a high-density electrode array (pixel distance down to the cellular size of 10- 15µm), where each individual electrode is coupled with a glutamatergic neuron to create a tight neuron-electrode coupling.  Following implantation of the hybrid prosthesis, the

glutamatergic neurons integrate and synapse with the host retinal circuits. Patterned electrical stimulation of these glutamatergic neurons by the electrodes modulates glutamate release onto the synapse with the host bipolar cells after which the remaining retinal circuitry is activated in an identical manner to natural vision. The ultimate electrode-neurons proximity and the low charge neural activation threshold allow for the significant reduction in electrode dimensions and an increase in pixel density as well as the continuous graded potential activation, thus mimicking the graded potential fashion of the bipolar and photoreceptor cells. Moreover, the indirect activation of the host retina by glutamatergic neurons (rather than direct electrical activation), can potentially preserve the natural visual circuits (e.g. ON and OFF).
In this talk I will present our results with generation of photoreceptor precursors from hESC. I will present the functional characterization of these cells by path-clamp and calcium imaging techniques, which demonstrated that the intracellular calcium of the cells can be electrically modified. I will further discuss the many challenges and approaches taken in device fabrication and tissue engineering toward the development of hybrid retinal prosthesis.
.

ההרצאה תתקיים ביום ראשון 19.11.17, בשעה 14:00
בחדר 315, הבניין הרב תחומי, אוניברסיטת תל אביב

Speaker: Lior Fritz,

M.Sc. student under the supervision of Prof. David Burshtein

Wednesday, November 22^nd, 2017 at 15:00

Room 011, Kitot Bldg., Faculty of Engineering

Simplified End-to-End MMI Training and Voting for ASR

Abstract

A simplified speech recognition system that uses the maximum mutual information (MMI) criterion is considered. End-to-end training using gradient descent is suggested, similarly to the training of connectionist temporal classification (CTC). We use an MMI criterion with a simple language model in the training stage, and a standard HMM decoder. Our method compares favorably to CTC in terms of performance, robustness, decoding time, disk footprint and quality of alignments. The good alignments enable the use of a straightforward ensemble method, obtained by simply averaging the predictions of several neural network models, that were trained separately end-to-end. The ensemble method yields a considerable reduction in the word error rate.

School of Mechanical Engineering Seminar
Wednesday, December 27, 2017 at 14:00
Wolfson Building of Mechanical Engineering, Room 206

Hydroelastic waves and their interaction with structures

Prof. Alexander Korobkin

(University of East Anglia, UK)

Co-author: S. Malenica (BV), T. Khabakhpasheva (Lavrentyev Institute of Hydrodynamics)

Linear problems of hydroelastic wave diffraction by structures with vertical walls are studied for a circular cylinder frozen in ice cover of constant thickness and infinite extent. The water depth is constant. The ice plate is modelled by a thin elastic plate clamped to the surface of the cylinder. The cylinder is mounted at the sea bottom. One-dimensional incident hydroelastic wave of small amplitude propagates towards the cylinder and is diffracted on the cylinder.  Deflection of the ice plate and the bending stresses in it are determined by two methods: (a) using the integral Weber transform in radial direction, (b) using the vertical modes for the fluid of constant depth with the rigid bottom and elastic upper boundary. The solution by the second method is straightforward but we cannot prove that the solution is complete because the properties of the vertical modes are not known. The solution by the Weber transform is more complicated but this solution is unique. We will show that these two solutions are identical. This result justifies the method of the vertical modes in the hydroelastic wave diffraction problems. For a circular cylinder the vertical-mode solution can be also justified by substitution. Different conditions at the contact line between the cylinder and the ice sheet are considered. The wave diffraction problem for broken ice is also considered. It is shown how the problem can be generalised to non-circular cylinders and interaction of several cylinders in ice.

‏

School of Mechanical Engineering Seminar
Monday, November 27, 2017 at 14:00
Wolfson Building of Mechanical Engineering, Room 206

An afternoon talk on femurs, arteries and high order finite element methods

Prof. Zohar Yosibash

School of Mechanical Engineering, Tel Aviv University

Predicting the mechanical response of human femurs and arteries is of major clinical importanceare for diagnostic purposes as well as for patient-specific treatment. Such prediction capabilities are being described which involve the combination of high order finite element (FE) methods, medical imaging and experimental observations.

Our research on human femurs will first be addressed, where we use quantitative computerized tomography (qCT scans) that enables a computer realization of subject-specific bone models. Combining qCT scans with high order FE simulations and a large set of experiments provide the opportunity to accurately simulate patient-specific femur’s response, allowing orthopedic surgeons to plan a proper patient specific treatment. Examples of the use of our methods in clinical practice for treatment of metastatic bone tumors will be provided.

If time allows, experimental and numerical methods for the analysis of human arteries will be addressed. Arteries are complex anisotropic structures undergoing large deformations and strains and are subject to active response due to contraction of smooth muscle cells. Attempts to properly describe their response will be described.

About the Lecturer

Prof. Yosibash received his B.Sc. in Aeronautical Engineering from the Technion (87), his M.Sc. in Applied Mathematics from Tel-Aviv Unviersity (92) and D.Sc. in Mechanical Engineering from Washington University, St. Louis, USA (94). He joined Ben-Gurion University in 1995 and since 2008 he is a full professor of mechanical engineering. He has been a visiting professor at Brown Univ from 2002-2007, and at the Technical Univ of Munich during 2010-2011. He received the Toronto prize for excellence in research at BGU in 2009. Prof. Yosibash was a Hans Fischer Senior Fellow at the Institute for Advanced Study at the Technical University of Munich (2009-2012) and serves as the scientific ambassador since 2013.  Prof. Yosibash joined the School of Mechanical Engineering at TAU in Oct 2017. He is the head of the lab for computational mechanics and experimental biomechanics at TAU and since 2015 serves as the president of the Israel Association for Computational Methods in Mechanics.