Karthikeyan Balasubramanian,
Doctoral Student,
Neural Instrumentation Lab,
Temple University.
==========================================================================
Academic Profile
M.Tech, Indian Institute of Technology – Delhi, 2004
B.Tech, Bharathidasan University – Trichy, 2002
Diploma, Institute of Textile Technology – Chennai, 1999
Current Research
Fiber based electrodes for intercortical recordings
One of the key components of BMIs (Brain Machine Interfaces) are a suitable electrode. The present research is to develop a conductive polymer based nanoscale electrode with optimized physical, electrical and mechanical properties. Polymer based electrodes have been shown to have better bio-compatibility than metal electrodes, in case of neural implants.
Real-time analysis of neural signals for information retrieval
Field Programmable Gate Arrays (FPGAs) are being used in the development process of an embedded system to analyze key neural signal processings, viz., spike detection and spike sorting.
Research Trails
Design and development of chemical protective Battle Dress Uniform (BDU) for soldiers using Artificial Intelligence Techniques
While functionality is always an essential property of BDUs, clothing comfort is gradually migrating from the segment of desired property to essential property. Functional properties are essentially incorporated by chemical finishes and their modes of applications. However, comfort properties involve a large number of variables, whose influence on perceive clothing comfort is not known precisely. In general, overall clothing comfort is influenced by its perceived comfort components viz., Tactile comfort, Thermal comfort and Pressure-fit comfort. Model free algorithms based on Artificial Neural Networks (ANNs) and Fuzzy Inference Systems (FISs) were developed to analyze the underlying relationship between material properties and sensory perceptions. In the process of reverse engineering, a hybrid system of Genetic Algorithm and ANN was put to use.
Scent-Infused Textiles
Melt-spun synthetic fibers have the potential to hold chemicals and gradually release them upon demanding conditions. Sheath-core and Sea-island fiber structures were developed with polypropylene (PP) as cores infused with scent and bio-degradable polylactic acid as sheath. Infusion of scent was a trial to evaluate the suitability of fibers to hold therapeutic chemicals.
Structural analysis of spunlaced and meltblown nonwovens using Digital Image Processing (Graduate Thesis)
Mechanical properties of nonwoven materials are greatly influenced by the structural arrangement of its constituent fiber matrix. This research was to develop a system to characterize Orientation Distribution Function (ODF), Diameter Distribution Function (DDF), etc. of spunlaced and meltblown nonwovens. Digital image processing technique was employed to determine the above-mentioned structural characteristics. The algorithm employs Hough transform of the polar kind for the orientation measurement and Medial Axis Transform in association with Distance transform for the diameter measurement. An attempt to predict the tensile properties of these fabrics, using the existing mathematical models and the objective values determined as an outcome of the measurement, was also proposed.
Machine Vision-based objective distortion measurement system for the evaluation of Canopy fabrics
Subjective evaluation of fabric defects on parachute canopies were mimicked via a vision-based objective distortion measurement system. Fast Fourier Transform coupled adaptive system was used to process the image to estimate the level of distortion after isolating the weave effects present in the image. Utility software named “Canopy Pro” was developed and is extensively used in the ADRDE (Aerial Delivery Research and Development Establishment) at Agra, India for objective evaluation of canopy fabrics’ distortion.
Development of an ANN embedded expert system for the design of canopy fabrics
An expert system was designed and developed to handle the Forward and Reverse engineering processes of canopy fabric development. The work describes the method of applying artificial neural network for the prediction of both construction and performance parameters for canopy fabrics. The design of the expert system embeds with it an ANN module for the property and parameter prediction. A utility software package, “CanEx”, was designed using the modules developed.
Publications
- Behera, B.K. and Karthikeyan, B, Artificial neural network embedded expert system for the design of Canopy fabrics, Journal of Industrial Textiles, Vol. 36, No. 2, 111-123 (2006)
- Karthikeyan, B and Sztandera, L, Sensory Perception of Fabrics: Analysis of Mechanical, HandFeel and Tactile Comfort Properties using Statistical and Artificial Intelligence Techniques, IASTED Conference Proceedings on Intelligent Systems and Control (Nov 2007)
- Yan Liu, Fernando Tovia, Karthikeyan, B, John D. Pierce, Jr, and Jeff Dugan, Scent Infused Textiles to Enhance Consumer Experiences, Journal of Industrial Textiles, vol. 37, No.1, 263 – 274 (2008)
- Karthikeyan, B and Obeid, I, Real time analysis of neural signals for information retrieval, 34th Annual Northeast Bioengineering Conference Proceedings, Brown University , 19-20 (Mar 2008)
- Karthikeyan, B and Sztandera, L, Analysis of Tactile Perceptions of Textile Materials using Artificial Intelligence Techniques – Part -1: Forward Engineering (accepted for publication, March 2009)
- Karthikeyan, B and Sztandera, L, Analysis of Tactile Perceptions of Textile Materials using Artificial Intelligence Techniques – Part -2: Reverse Engineering using Genetic Algorithm Coupled Neural Network (accepted for publication, March 2009)
Societies
- Student Member of Biomedical Engineering Society
- Student Member of the IEEE
- World Community Service Center – Center for Simplified Kundalini Yoga
