Posters

P1. A Computational Model to understand Hemispheric Specialization in Motor Control

1 Sundari Elango; 2 Prof. V. Srinivasa Chakravarthy; 3 Dr. Pratik Mutha 

1. Dept. of Biotechnology, Indian Institute of Technology Madras, Chennai 600036, Tamil Nadu, India; 2. Biological Engineering & Cognitive and Brain Sciences, Indian Institute of Technology, Gandhinagar 

Abstract:  Motor adaptation studies have shown transfer of motor memories from one arm to the other. In visuomotor rotation adaptation task, when Directional Error (DE) at peak velocity was used as performance measure, right arm showed better performance when subjects previously learnt the same task with the left arm, compared to the naive arm condition. But the reverse was not true, i.e., left arm showed no such advantage, when evaluated in terms of DE. However, when performance measure was changed from DE to final end position error left arm showed better performance when subjects had previously learnt the same task with the right arm, compared to the left arm under naïve condition. Thus, the left hemisphere, shows specialization for learning direction of movement, while the right hemisphere, shows specialization for learning distance of movement. Based on these results, we propose a simple motor control scheme distributed over the two brain hemispheres. Here, movement control is divided into two parts – direction and distance. In order to check the efficiency of the scheme in controlling movement, it is implemented in a neural network framework used to replicate certain experimental results related to visuomotor rotation adaptation.

P3. Virtual Reality-Enhanced Motor Imagery Brain-Computer Interfaces for Stroke Rehabilitation

 1.Saher Soni;  2.Sundari Elango;  3.V.  Srinivasa  Chakravarthy

1. IIT Gandhinagar; 2, 3 CNS Lab,  Department of Biotechnology, IIT Madras

Abstract:  Virtual reality (VR) has emerged as a promising technology in augmenting motor function recovery in stroke rehabilitation by offering novel strategies through immersive and customizable environments. The principle of Motor Imagery Brain-Computer Interfaces (MI-BCIs) lies in harnessing the brain's capacity to simulate movements mentally, enabling users to control external devices through the decoded motor-related brain activity. In this study, we investigated the synergistic potential of integrating VR and MI-BCIs as a comprehensive rehabilitation strategy tailored for stroke patients with upper limb motor disabilities. We developed a Unity-based 2D game within the Unity software framework, aimed at enhancing upper limb motor abilities and cognitive capacities such as attention and working memory. Furthermore, we constructed a classifier network, utilizing an open-source BMIdataset, to discern grasping hand movements by analyzing Electroencephalography signals. Our study showcases the successful design and implementation of these interventions, offering their potential utility for augmenting stroke rehabilitation.

P5. Comprehensive data and metadata management in a multi-lab collaboration through integration of open-source solutions

Reema Gupta;Thomas Wachtler

Ludwig-Maximilians-Universität München

Abstract: 

Recent technological advancements in electrophysiology facilitate the simultaneous recording of hundreds of channels while capturing complex behaviours. However, these extensive and complex datasets often remain underutilized due to insufficient solutions for data management and sharing. We address these challenges within the multi-lab collaboration In2PrimateBrains [1], an EU-funded international training network to investigate brain networks in the non-human primate.  We utilize a modular and integrated set of open-source tools, methods, and services to achieve user-friendly solutions that balance standardization with adaptability.

Metadata collection is facilitated by the lightweight odML [2] metadata format, enabling automated data annotation and context provision in a human- and machine-readable manner [3]. Standardization is achieved by utilizing a set of controlled terminologies and metadata schemas built upon existing community efforts, including BIDS extension BEP032 [4] and the EBRAINS openMINDS [5].

We utilize Neo [6] to achieve a common data representation and as an I/O bridge to interface with the variety of data formats used in participating labs. To ensure data accessibility, we employ the NIX data format [7] for storage, which enables comprehensive organization and integration of metadata alongside various types of data.

For data sharing, we use GIN [8], a collaborative data platform. GIN offers version control for research data, fine-grained access control, collaborative features, and data publication services.

Standardized representation of data in open formats, along with comprehensive metadata, and programmatic access ensure that the data is interoperable for further processing by generic analysis scripts, tools, and services [9,10,11]. In conclusion, we present a case study of heterogeneous data and metadata management in a multi-lab collaboration. Our approach is to employ and build upon an ecosystem of open tools, methods, and services that collectively work towards making research data more FAIR [12].

Acknowledgements:

Supported by the European Union’s Horizon 2020 research and innovation programme (Grant agreement No 956669)  

P7. Molecular Docking Investigation of TREM2 against a family of Amyloid-β (Aβ) ligands with Therapeutic potential against Alzheimer’s disease 

SK Chand Basha ; Mekala Janaki Ramaiah ; Jagannatha Rao Kosagisharaf 

Department of Bio-Technology, Koneru Lakshmaiah Education Foundation, Vaddeswaram, AP, India 

Abstract: Alzheimer’s disease (AD) is a major cause of dementia, which severely impairs the overall personality traits of the effected individual, besides cognitive impairment. In the light of current knowledge, which indicates the potential role of Microglial Triggering receptor expressed on myeloid cells 2 (TREM2) in the amelioration of amyloid-β (Aβ), we propose a hypothesis on TREM2 pathway. To examine our hypothesis we have conducted the current In silico Molecular Docking studies. Aβ are considered as the chief pathophysiological attribute of AD, diminishing of the same is a challenge. TREM2 3D structure has been predicted by using I-TASSER server. ClusPro server is utilised for conducting molecular docking studies. Totally, 6 different variants of Aβ ligands are used: Aβ 6, Aβ 40, Aβ 42, Aβ 42A, Aβ 42B and Aβ oligomer. The coordinates of the Aβ ligands are imported from PDB database. In ClusPro, ranking of docking models are based on the cluster size, and the results suggests the number of clusters for top docking models of respective ligands are Aβ 6 – 337; Aβ 40 – 154; Aβ 42 – 125; Aβ 42A – 108; Aβ42B – 121 and Aβ oligomer – 101. Therefore, the results suggests that TREM2 has highest affinity for Aβ6 ligand and have potential therapeutic value. 

Keywords: Alzheimer’s disease, TREM2 and Aβ. 

P9. A comprehensive Gaming System for Rehabilitation of Upper Extremity Post Stroke

S. Hafsath1, S. Elango1, A.J.A. Francis2, D. Darshini1, V.S. Chakravarthy1, P.N. Sylaja3, R. Datsen

1. Indian Institute of Technology, Department of Biotechnology, Chennai, India, 2. Indian Institute of Technology, Department of Cognitive Science, Kanpur, India, 3. Sree Chitra Tirunal Institute for Medical Sciences and Technology, Neurology, Thiruvananthapuram, India

Abstract: Hemiparesis of the upper limb is the most common disability after stroke. Physiotherapy while, crucial for recovery, is repetitive in nature and cause patients to get bored and lose motivation to continue therapy. Recently, use of Virtual Reality (VR) has been advocated to promote therapy. With VR, movements used in therapy are gamified, thus providing a context to them, and motivating patients to move. Here, we have developed a virtual gaming environment containing 5 games. The system is also capable of tracking a patient's progress via a tracking module. The tracking module tracks movements made using a webcam and translates them into game events. The gestures were chosen based on their relevance to Activities of Daily Living (ADL) and includes abduction/adduction of shoulder, flexion/extension of elbow and fingers, and pronation/supination of wrist. The system is currently undergoing clinical trials involving stroke patients with hemiparesis. The main aim of the study is to evaluate whether the proposed system can work better than conventional therapy at promoting recovery.

P11. Impact of Perturbational Landscape of a Power-coupled Oscillator Network on Attributes of Complex Brain Graph Network

ANIRBAN BANDYOPADHYAY(1); SAYAN GHOSH (1); DIPAYAN BISWAS (1); BAPI RAJU S (2);  V. SRINIVASA CHAKRAVARTHY (1)

IIT Madras (1); IIIT Hyderabad (2)

bstract:  Power-coupled Hopf-oscillator system-based complex network has been efficacious in reconstructing neuroimaging signals like- BOLD and Functional Connectivity (FC) from structural Connectivity (SC). In silico perturbation like proportional thresholding causes the anomalous behaviour of FC owing to pruning on SC. It is observed that the parameters of the Hopf oscillator network can restore the FC, which is distorted due to alteration in SC. Similarly, graph network properties of the brain network are restored during the in-silico rehabilitation process, where degree distribution, small worldness, clustering coefficient, participation coefficient, modularity and community structure, and assortativity has been estimated and compared for different perturbational landscape. Such work will pave the path for understanding neurological disorders like schizophrenia and stroke, where the global functional organization of the brain network is distorted due to structural disruption, and the parameter-based restoration process can be fruitful for predicting possible treatment procedures by brain-therapeutic technologies for those neurological disorders.   

P13. Simple microprocessor model inspired by Hippocampus

Vigneswaran C; V. Srinivasa Chakravarthy

1. CNS Lab,  Department of Biotechnology, IIT Madras

Abstract: 

Neural Turing Machines (NTMs) perform composite operations of I/O, arithmetic, and memory functions concurrently with an input instruction. Every component in NTMs is differentiable, and memory operations over the elements are realized in the neuronal activations. However, traditional NTMs have an explicit memory array to access the items to perform executive functions, and it significantly reduces the biological feasibility of the model. Our model is inspired by the hippocampus and its attributed index theory to perform basic microprocessor operations. The model can emulate simple I/O, arithmetic, and memory functions where all the operations are expressed in the network dynamics. 

P15. Graph theory application with functional connectivity to distinguish early mild cognitively impaired and mild cognitive impaired patients from healthy controls.

Aditya Balkrishna Umarjikar ; Stephen Mayhew

CNS Lab, Department of Biotechnology, IIT Madras

Abstract:  Inividuals with Alzheimer's Disease (AD) experience significant structural and functional brain alterations. Limited studies focus on Early Mild Cognitive Impairment (EMCI) and Mild Cognitive Impairment (MCI), which have a high probability of progressing to AD. We used resting-state fMRI data from EMCI, MCI, and Cognitively Normal (CN) subjects aged 60-69 from the Alzheimer's Disease Neuroimaging Initiative (ADNI) to model the brain as a graph-based network using 53 selected regions of interest (ROIs) involved in memory and cognitive functions. The functional connectivity (FC) analysis across 164 ROIs revealed no significant differences between CN and MCI subjects. However, EMCI subjects showed increased FC compared to MCI, and CN subjects had higher FC than EMCI subjects. Theoretical graph analysis showed decreased nodal centrality in MCI subjects compared to CN subjects, while an increase in nodal centrality was observed in areas of the default mode network. These findings highlight the potential of theoretical graph analysis in understanding brain region characteristics and network complexity.