Srivathsa Pasumarthi

AI in Radiology Machine Learning Engineer South Indian Classical Musician

About

I am a Machine Learning engineer currently applying novel deep learning techniques on radiological datasets to achieve faster scans and lesser radiation. A former graduate research assistant at the Johns Hopkins Medical Institute, I have worked on various cardiac MR and CT datasets during my term at the Johns Hopkins University where I received my masters in Computer Science. Within the field, the applications of deep learning in Computational Genomics, Computational Neuroscience and Medical Imaging are my special interests. On a personal front, I am also a South Indian Classical percussionist and a flautist.

View Résumé

Skills

Machine Learning

Cloud

Python

Keras

PyTorch

Caffe

Tensorflow

AWS

GCP

Heroku

Web

Databases

Node.js

React

Angular

HTML5

CSS3

MySQL

PostgreSQL

mongoDB

Others

DICOM

Java

Docker

Linux

Python

Keras

PyTorch

Caffe

Tensorflow

AWS

GCP

Heroku

Node.js

React

Angular

HTML5

CSS3

MySQL

PostgreSQL

mongoDB

DICOM

Java

Docker

Linux

Education

MSE Computer Science

August 2017 - May 2019

Johns Hopkins Univeristy, Baltimore, MD

Coursework

Machine Learning - A Probabilistic Perspective
Deep Learning
Computational Genomics: Sequences
Computational Genomics: Applied Comparative Genomics
NeuroData Design
Probabilistic Models of the Visual Cortex
Data Intensive Computing

BE Computer Science & Engineering

August 2010 - 2014

Anna University, Chennai, India

Project

Android application to predict the arrival time of public transport buses using GPS and Data Mining.

Work Experience

Machine Learning Engineer

July 2019 - Present

Subtle Medical Inc., Menlo Park, CA.

Applying novel deep learning techniques to enhance radiological images for achieving faster scans and less radiation.

Graduate Research Assistant

January 2018 - May 2019

Lima Lab, Johns Hopkins Medical Institute, Baltimore, MD.

Applying deep learning on Cardiac MR and CT studies to automate ventricle segmentation, calcium scoring and scar segmentation using Keras.

Machine Learning Intern

May - August 2018

Arterys, San Francisco, CA.

Drove the automation of an entire cardiac radiology workflow in a scaled agile framework, starting from data collection, data munging, Keras based 2D segmentation modeling using Cardiac MR data and also co-ordinated with clinicians for clinical usability assessment.

Data Science Intern

May - August 2017

Ajira Tech, Chennai, India.

Used hierarchical clustering and collaborative filtering on large scale retail data for a retail marketing and customer engagement software, to learn propensity, purchase patterns and demography.

R & D Engineer, Machine Learning

July 2016 - May 2017

Qube Cinema Technologies, Chennai, India.

Prototyped a CNN based image classifier using Caffe Python to count the number of people in a movie auditorium, in order to provide real time viewership information to ad providers.

Full Stack Web Developer

May 2014 - July 2016

Qube Cinema Technologies, Chennai, India.

Developed many significant features over the full stack of QubeWire - a cloud based digital cinema distribution management system, using Node.js and React.JS and PostgreSQL, in an agile setting.

Projects

Music

I am a South Indian classical percussionist with over 19 years of training and appearances in many regional and national TV channels. I am also a recognized musician of All India Radio (a subsidiary of Government of India) and performed in over 600 concerts in India and in the US, including the popular Cleveland Thyagaraja Aradhana. I am also a Indian Classical Flautist and lead a fusion band called Yuvadhwani.

Below are a few videos from my live performances, TV telecasts and my fusion band.