Areas of Interest

Video/Image/Signal Processing - Biomedical Applications

Current Interests

  • Video Processing: Non-contact monitoring of vitals from animal models (Zebrafish larvae),

Estimation of heart-rate from face-videos

In vitro assessment of the Physical Response of Blood vessels using microscope-videos,

Flow-analysis in Nuclear Medicine

  • Digital Pathology & Microbiology: TB Diagnosis (hold a US Patent in this area), Candida-image Processing

  • Development of Signal Processing Algorithms: Automatic QRS-detection, adaptive signal processing algorithms for addressing signals with time-varying characteristics, Recording and processing Naadi Signals towards assessing physiological conditions of a subject.

  • Computed Tomography - Algorithms for Image Reconstruction/Enhancement, Re-projection

v_zf_website.mp4

We can estimate the heart-rate of zebrafish Larvae completely automatically, from microscope videos

Zebrafish is recognized as an important animal model for assessing the cardiovascular system

Collaborator: Dr. Kiranam Chatti, DRILS, Hyderabad

RRG_Naadi_Picture-1.pptx

Naadi signals: Signals tapped from the three naadi spots on the wrist: Vaatha (V), Pittha (P) and Kapha (K)


RRG_Naadi_Picture-2.pptx

The Naadi-acquisition System

Collaborator: Dr. Arul Amuthan

Siddha Division, Manipal-Melaka Medical College, MAHE< Manipal - 576104.

RRG_Naadi_Acq_Demo.mp4

Demo

Real-time acquisition of V,P & K waveforms.

Shown above, is an example of real-time acquisition of V,P & K waveform.

K1- Sample Video Clip.avi

A 30s video of the face, suffering fluctuations in illumination, and corrupted by facial expressions.


James J., S. Krishna and RR Galigekere, "Automatic and Adaptive Signal- and Background-ROIs with Analytic-Representation-based Processing for Robust Heart-rate Estimation by a Webcam ", IEEE Access https://ieeexplore.ieee.org/document/8998210?source=authoralert.

We can count the number of bacilli in ZN-stained sputum smears automatically, irrespective of their shape, size, and the presence of staining artifacts.

This assists the technician: manual counting - as per protocol - is tedious and error-prone.

Collaborator: Dr. Vishnu Kumar, Dept. of Microbiology & Mr. Rijul Soans, Dept. of Electrical Engineering, IIT Delhi

We hold a US Patent in this area (Details in 'Publications' Section)

We can estimate an image with a region of interest in Computed tomography (CT) from truncated projections, by simple methods of extrapolating the truncated projections.

Gaurav Varma & RR Galigekere, ICVGIP 2010, IIT Madras, Chennai.RR Galigekere & DW Holdsworth, LNCS 2879 (Springer), pp. 126-133, 2003. ( MICCAI 2003, Montreal, Canada)