Skip to content
View pranay1990's full-sized avatar

Block or report pranay1990

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
pranay1990/README.md

Hi, I'm Pranay Barkataki πŸ‘‹

AI/ML Researcher Β· Medical Imaging & Neurovascular AI Β· PhD in Physics
Building deep learning systems that help clinicians detect strokes faster.

LinkedIn Google Scholar Email PyPI


About Me

I'm an AI/ML Researcher at Nuvo AI, where I'm a core contributor to NeuralIQ β€” a clinical AI platform for stroke neurovascular care.

My work sits at the intersection of deep learning and high-stakes clinical decision-making: building, validating, and deploying AI pipelines that operate directly in hospital reading rooms, helping radiologists detect hemorrhages, identify vessel occlusions, and decide on treatment in minutes rather than hours.

I hold a PhD in Physics from VNIT Nagpur, which gives me an unusual edge: I approach ML problems with the rigour of a scientist and the pragmatism of an engineer.


🧠 What I Work On

NeuralIQ β€” Clinical Stroke AI (Nuvo AI Β· Jan 2025–Present)

Module What I built Result
NCCT Hemorrhage Segmentation 3D U-Net pipeline trained on 3,000+ patient studies DICE 0.73 (hemorrhagic), 0.58 (ischemic)
LVO Detection on CTA Deep learning pipeline for Large Vessel Occlusion detection AUC 0.92 Β· Recall 0.87 β€” Published: Medical Image Analysis, Elsevier 2025
Multi-modal Pipeline Unified NCCT + CTA stroke quantification 44% reduction in door-to-treatment time
Clinical Decision Support ASPECTS scoring, tPA/MT eligibility, midline shift Deployed live in 47+ hospitals

πŸ“‹ Filed IP: 3D Deep Learning-Based Stroke Segmentation and Volumetric Analysis Pipeline for Clinical CT Imaging


πŸ”¬ Research & Publications

  • 7 peer-reviewed journal articles across Physics, AI, and Computational Science
  • 1 book published
  • 1 patent application filed (medical AI)

πŸ‘‰ Full list: Google Scholar Profile


πŸ› οΈ Tech Stack

skills = {
    "Languages":       ["Python", "C++", "FORTRAN", "JavaScript"],
    "DL Frameworks":   ["PyTorch", "TensorFlow", "NNabla", "Hugging Face"],
    "Medical Imaging": ["3D U-Net", "DICOM", "SimpleITK", "OpenCV",
                        "NCCT/CTA/CTP pipelines", "Volumetric segmentation"],
    "ML / Data":       ["XGBoost", "CatBoost", "ARIMA", "PySpark",
                        "Databricks", "Scikit-learn", "Pandas", "NumPy"],
    "Generative AI":   ["Diffusion Models", "Consistency Trajectory Models",
                        "LLM Fine-Tuning", "GRPO Reinforcement Learning"],
    "Domains":         ["Computer Vision", "Medical Image Analysis", "NLP",
                        "Neurovascular AI", "Statistical Modelling"],
}

πŸ“¦ Open Source

QuantumInformation β€” Python Library

A Python library for quantum information and computation.

pip install QuantumInformation

πŸ† Highlights

  • πŸ“„ Paper accepted in Medical Image Analysis (Elsevier, 2025) β€” IF 10.7
  • πŸ§ͺ Trained diffusion models on 8Γ— NVIDIA H100 GPUs β€” FID 2.37 (state-of-the-art)
  • πŸ₯‡ Best PoC Team Award β€” Sony
  • πŸŽ“ PhD in Physics β€” VNIT Nagpur

πŸŽ“ Education

Degree Institution Year
PhD β€” Physics (Computational / Mathematical Modelling) VNIT Nagpur 2015–2021
M.Sc. β€” Physics VNIT Nagpur 2013–2015
B.Sc. β€” Physics Hansraj College, University of Delhi 2009–2012

πŸ“¬ Get In Touch

I'm always open to conversations about medical AI, neurovascular applications, or anything at the intersection of rigorous science and engineering.

πŸ“§ pranay.barkataki@gmail.com πŸ”— linkedin.com/in/pranaybarkataki


Profile views

Pinned Loading

  1. G2NET_Gravitational_Wave_Detection G2NET_Gravitational_Wave_Detection Public

    Kaggle compitetion to detect Gravitational Wave

    Jupyter Notebook

  2. US_Patent_Phrase_Matching US_Patent_Phrase_Matching Public

    Jupyter Notebook

  3. Udacity_content_based_recommendation Udacity_content_based_recommendation Public

    Jupyter Notebook

  4. Qiskit/qiskit Qiskit/qiskit Public

    Qiskit is an open-source SDK for working with quantum computers at the level of extended quantum circuits, operators, and primitives.

    Python 7.6k 3k

  5. Deforestation_SQL_Udacity Deforestation_SQL_Udacity Public

    This is the project in SQL nanodegree program from Udacity

    1 1