This project employs EEG signal analysis to classify sleep stages using a low-computation neural architecture and the fewest EEG channels possible. The aim of the project was to build an assistive tool for manual sleep scorers. This 5 stage sleep classifier can provide information on sleep quality and assist in diagnosing sleep disorders. 

The confusion matrix on the right is based on one of the initial models trained in this project.

This project focused on identifying the most suitable sensor combination for developing a physical activity monitor. The selection criteria include reliability, sensitivity to external noise, power consumption, computational time, and accuracy. 

It contains methods to remove the effect of gravity on accelerometer data, and detect peak using MATLAB functions.

Simulated an antenna design for a non-invasive glucose monitor. Later as part of this project tested the physical attributes of this antenna design through a microfabrication lab. Build electronic circuits and algorithms to generate pulses and capture, filter, and amplify the received signals to detect changes in glucose concentration. 

Software used Ansys HFSS and Zeland IE3D to design and test both the directive and the microstrip antenna's design. The images on the right are a few prototype designs.

Co founded a venture 'Espy'. Designed and integrated QR codes on t-shirt design, enhancing the functionality of wearable clothing.  Enhanced the interactivity through the implementation of Augmented Reality (AR) using Layar. This initiative transformed conventional t-shirt designs into dynamic, interactive experiences. We catered to designs for college fests, college societies, NGOs and individuals. 

The video on the right was made for the Enactus KISS-KIIT chapter.

A Python script to facilitate the notification and auto booking of vaccine slots through the Co-WIN Public APIs. This script checks available slots in vaccination centres at the areas of interest (by district or pincode) and notifies users as soon as slots become available (using a beep sound, winsound package) and try to book the slots for the selected beneficiaries.

github.com/bikramkhastgir/COWIN_COVID_vaccine_slot_autobook 

This project explores the use of mRNA analysis to trace common ancestry and interspecies breeding events between multiple mammals. 

By constructing phylogenetic trees and examining genetic alignments, the study aims to determine whether Peto's paradox, observed in large-bodied mammals like whales and elephants as well as in small-bodied animals like turtles and mice, arose independently or from a shared ancestor, shedding light on evolutionary biology.

The pictures on the right depict one such analysis between extinct Mammoths, elephants and the blue whale. 

Built a mathematical model of a synthetic gene network, focusing on analysing genetic oscillators as outlined by Hasty et al. (2002). 

The study employs sophisticated simulation techniques done using XPPAUT to explore the dynamics of synthetic gene oscillators. It models the complex interplay of genetic components and their implications in cellular processes. This intersection of theoretical biology and applied mathematics offers insights into the role of self promoters and suppressors to recurrently control the expression of individual genes and their oscillatory behaviours.

Control Technology for EEG Based feedback loop to control motor feedback to develop sensitive touch control for prosthetics.