Alexei Ossadtchi is a leading Russian scientist in the field of Neuroimaging. He is currently a Professor and Director of the Center for Bioelectric Interfaces at the National Research University Higher School of Economics. Dr. Ossadtchi holds a Ph.D. degree in Electrical Engineering from the University of Southern California (2003), where he developed approaches for automatic identification of epileptogenic regions in multi-focal epilepsy. These methods are being applied to non-invasively collected magnetoencephalographic (MEG) data for improved diagnostic accuracy of patients with epilepsy. Dr. Ossadtchi continued this research in Russia and developed efficient solutions to the interpretation of MEG and electroencephalography (EEG) based neuroimaging in the clinical context. His recent developments include algorithmic techniques for solving the EEG and MEG inverse problem on the group level, novel robust analytical solutions improving spatial resolution of MEG based neuroimaging and providing access to detection of functional coupling of neuronal sources. His software algorithms were applied to the analysis of brain activity in `mathematical methods in multimodal neuroimaging. Dr. Ossadtchi authored more than thirty publications in high-impact international journals specialized in methods for functional neuroimaging and their real-time applications, such as BCIs and neurofeedback. In addition to academic background, Dr. Ossadtchi has an extensive business experience. He was a Senior Scientist for twelve years at Source Signal Imaging Inc., San Diego, CA, where he developed the EMSE Software Suite, a versatile data analysis tool for multimodal neuroimaging.
Talk 1: Magnetoencephalography: from sensor data to source distributions with conventional and advanced methods
Magnetoencephalography (MEG) is a functional brain imaging modality that allows us to visualize brain function at the millisecond temporal scale with subcentimeter spatial resolution. This fantastic property of MEG results from the hardware and algorithmic advances implemented in the modern MEG systems and highly specialized data-processing software. In order to translate MEG sensor measurements into cortical activation maps we solve the inverse problem (IP) that appears to be ill-posed. Methods for circumventing this undesired property play a pivotal role in defining the resultant spatial resolution of the entire MEG imaging modality.
In my talk I will first introduce the basic mathematics behind solving the inverse problem of MEG (and EEG). Then, I will describe a family of methods for solving the IP developed in my lab that yield an improved spatial resolution and allow for imaging functional connectivity between cortical sources. I will conclude with a brief description of the software for real-time MEG and EEG based functional imaging developed by us that solves the inverse problem on the fly and operates at the speed of the brain on a regular PC!
Talk 2: Various disguises of motor-imagery BCI
En: In this informal talk\demonstration we will describe the basics of motor-imagery brain-computer interface technology. In parallel we will give you an opportunity to try using such BCI yourself.
Ru: Мы начнем лекцию с обзора основных типов неинвазивных нейроинтерфейсов. Тех, которые можно попробовать без необходимости вживления электродов в кору головного мозга. Я расскажу о том, что вы никогда не услышите от продавцов и большинства разработчиков, убеждающих вас, что силой мысли при помощи одного электрода на лбу вы сможете управлять дроном, используя при этом активность вашего мозга. Вы также узнаете о том, что большинство из парадигм неинвазивных нейроинтерфейсов не имеют ничего общего с чтением ваших мыслей. Они вообще не про мысли и интенции, а скорее про реакцию мозга на внешнюю стимуляцию.
Особое внимание мы уделим самому привлекательному нейроинтерфейсу с точки зрения романтической идеи чтения мыслей и его применении в реабилитации постинсультных больных, а также больных c параплегией. Это идеомоторный нейроинтерфейс, который основан на воображении движений.
Как мне кажется, по-настоящему естественного управления можно достичь только за счёт инвазивных интерфейсов. Поэтому я закончу лекцию рассказом про нашу работу по созданию первого в России двунаправленного нейроинтерфейса на основе электрокортикографических сигналов.
Demonstrations
