Praveen Gurunath Bharathi

(Medical Image Analysis, Nailfold Capillaroscopy, MRI Post Processing, Deep Learning)
Postdoctoral Research Fellow, Nuclear Medicine and Molecular Imaging, Department of Radiology, Stanford Medicine

Email  /  Google Scholar  /  LinkedIn  /  Twitter

I am currently working as a Postdoctoral Research Fellow in Nuclear Medicine and Molecular Imaging, Department of Radiology, Stanford Medicine, supervised by Dr. Farshad Moradi , Dr. Mehdi Khalighi and Dr. Andrei Iagaru . I am working on building algorithms to predict and monitor real-time PET image quality using AI. Prior to this, I worked as a Postdoctoral Research Associate at the Division of Informatics, Imaging & Data Sciences, The University of Manchester, supervised by Prof. Chris Taylor, Prof. Ariane Herrick and Dr. Michael Berks. I developed an inexpensive imaging & automated analysis system to support early diagnosis of Systemic Sclerosis.

Previously, I completed my Ph.D. at the Department of Electrical and Electronics Engineering, Birla Institue of Technology and Science - Pilani (BITS Pilani), Goa, India, under the supervision of Prof. Anita Agrawal. My thesis was focused on "Development of an Automated Intelligent Decision Support System for Multiple Brain Disorder Diagnosis from MRI Scans". I have worked as a Project Intern with Prof. S. N. Omkar at Indian Institute of Science (IISc)-Bangalore. Before joining BITS Pilani, I obtained my Master's degree from R. V. College of Engineering, Bangalore and Bachelor's degree from Visvesvaraya Technological University, Belgaum, India.

My research is mainly at the interdisciplinary field of deep learning and medical image analysis. In particular, I am applying the deep learning models to improve disease prognosis.

[Updates]
  • [08/2023] Our paper titled "Noise and Sharpness estimation in Positron Emission Tomography (PET) images using radiomics and deep learning'' has been accepted for Top Rated Oral Presentation (TROP) at EANM 2023

  • [08/2023] Our paper titled "Elevating Amodal Segmentation using ASH-Net Architecture for Accurate Object Boundary Estimation'' has been published in IEEE acess

  • [06/2023] Our poster titled "Rapid dynamic reconstruction using list mode data for monitoring PET image quality accurately predicts final image noise and perceived quality" presented at 2023 SNMMI Annual Meeting

  • [01/2023] Our paper titled "A deep learning system for quantitative assessment of microvascular abnormalities in nailfold capillary images" has been accepted for publication in Rheumatology

  • [11/2022] Started my Postdoc at Nuclear Medicine and Molecular Imaging, Department of Radiology, Stanford Medicine

  • [09/2022] Delivered a talk on "Application of Data Science for Medical Image Analysis" at Cancer Systems Imaging, The University of Texas MD Anderson Cancer Center

  • [05/2022] Our paper titled "Nailfold capillaroscopy: a survey of current UK practice and ‘next steps’ to increase uptake amongst rheumatologists'' has been accepted for publication in Rheumatology

  • [04/2022] Oral presentation of our paper titled "Development of an automated deep learning-based system for distinguishing between systemic sclerosis and normal capillaries'' at British Society for Rheumatology (BSR)

  • [04/2022] Poster presentation of our paper titled "Nailfold capillaroscopy: a survey of current UK practice and ‘next steps’ to facilitate generalised uptake'' at British Society for Rheumatology (BSR)

  • [03/2022] Poster presentation of our paper titled "Development of an automated deep learning-based system for distinguishing between systemic sclerosis and normal capillaries'' at 7th Systemic Sclerosis World Congress

  • [08/2019] Successfully defended my PhD thesis

  • [07/2019] Started my Postdoc at Division of Informatics, Imaging & Data Sciences, The University of Manchester

  • [Research]
    A deep learning system for quantitative assessment of microvascular abnormalities in nailfold capillary images
    Rheumatology, 2023

    Praveen Gurunath Bharathi, Michael Berks, Ariane Herrick, Graham Dinsdale, Andrea Murray, Joanne Manning, Sarah Wilkinson, Chris Taylor

    We mimicked the image interpretation strategies of SSc experts, using deep learning networks to detect each capillary in the distal row of vessels and make morphological measurements. We combined measurements from multiple fingers to give a subject-level probability of SSc. Fully automated analysis using deep learning can achieve diagnostic performance at least as good as SSc experts, and is sufficiently robust to work with low-cost digital microscope images.

    Paper
    Nailfold capillaroscopy: a survey of current UK practice and ‘next steps’ to increase uptake amongst rheumatologists
    Rheumatology, 2023

    Martin Eden, Sarah Wilkinson, Andrea Murray, Praveen Gurunath Bharathi, Andy Vail, Chris Taylor, Katherine Payne, Ariane L Herrick

    An online survey was developed using expert opinion from clinicians, scientists and health service researchers. The survey was piloted and sent to UK-based rheumatologists using established electronic mailing lists between October 2020 and March 2021. Survey data were analysed using descriptive statistics.

    Paper
    P117 Nailfold capillaroscopy: a survey of current UK practice and ‘next steps’ to facilitate generalised uptake
    Rheumatology, 2022

    Martin Eden, Sarah Wilkinson, Andrea Murray, Praveen Gurunath Bharathi, Chris Taylor, Katherine Payne, Ariane L Herrick

    Substantial variation in approaches to the diagnosis of SSc across the UK was identified. Potential benefits of a standardised system were described by respondents including the improved diagnosis and management of SSc, realising potential patient benefits and reducing current health inequalities. Survey findings provide evidence to help develop future studies to develop and evaluate the proposed new system.

    Paper
    OA08 Development of an automated deep learning-based system for distinguishing between ‘systemic sclerosis' and ‘normal' capillaries
    Rheumatology, 2022

    Praveen Gurunath Bharathi, Michael Berks, Ariane Herrick, Graham Dinsdale, Andrea Murray, Joanne Manning, Sarah Wilkinson, Chris Taylor

    We have developed a completely automated nailfold capillary detection and analysis system based on deep learning. Apex detection, distal classification, and subsequent computation of distal apex measurements from nailfold capillaroscopy images are all performed with no human input. We have combined image-level features and have built a subject-level classifier model which successfully discriminates between SSc and HC/PRP with an improved AUC (i.e. improved sensitivity and specificity) compared to our previous work.

    Paper
    Combination of hand-crafted and unsupervised learned features for ischemic stroke lesion detection from Magnetic Resonance Images
    Biocybernetics and Biomedical Engineering, 2019

    Praveen Gurunath Bharathi, Anita Agrawal, Ponraj Sundaram, Sanjay Sardesai

    We present a novel automatic method to detect acute ischemic stroke lesions from Magnetic Resonance Image (MRI) volumes using textural and unsupervised learned features. The proposed method proficiently exploits the 3D contextual evidence using a patch-based approach, which extracts patches randomly from the input MR volumes.

    Paper
    Brain abnormality detection using template matching
    Bio-Algorithms and Med-Systems, 2018

    G.B. Praveen, Anita Agrawal, Shrey Pareek, Amalin Prince

    A technique for detecting abnormalities in brain MRI using template matching is proposed. Bias filed correction is performed on volumetric scans using N4ITK filter, followed by volumetric registration. Normalized cross-correlation template matching is used for image registration taking into account, the rotation and scaling operations. A template of abnormality is selected which is then matched in the volumetric scans, if found, the corresponding image is retrieved. Post-processing of the retrieved images is performed by the thresholding operation; the coordinates and area of the abnormality are reported.

    Paper
    Ischemic stroke lesion segmentation using stacked sparse autoencoder
    Computers in Biology and Medicine, 2018

    G.B. Praveen, Anita Agrawal, Ponraj Sundaram, Sanjay Sardesai

    We propose an unsupervised featured learning approach based on stacked sparse autoencoder (SSAE) framework for automatically learning the features for accurate segmentation of stroke lesions from brain MR images.

    Paper
    MediCloud: Cloud-Based Solution to Patient’s Medical Records
    Proceedings of the International Conference on ISMAC in Computational Vision and Bio-Engineering 2018 (ISMAC-CVB), 2018

    G.B. Praveen, Anita Agrawal, Jainam Shah, Amalin Prince

    We have proposed a public cloud-based “Infrastructure as a Service (IaaS)” model as a solution for maintaining the medical records of patients. New patient information is registered at the hospital with the help of a unique identification number. For each user, the bucket is created in the Amazon AWS cloud to store or retrieve the data. Data access is performed with the help of username and password provided by the web link which is embedded in the unique QR code. Two QR codes are used, the first code gives the access to the login page, whereas the latter one is used for accessing the corresponding user bucket.

    Paper
    An Analysis of Leg Muscle Stretch Using 3D Digital Image Correlation
    International Journal of Organizational and Collective Intelligence (IJOCI), 2017

    GB Praveen, S Raghavendra, Victor IC Chang

    The paper presents a generic methodology to compute the strain pattern in the Sural and calcaneal region during leg dorsiflexion experiment. In the experiment, the subject is made to stand on an inclination plane and images are captured at varying angular inclinations. Strain plots obtained after comparison indicates the strain distribution in the posterior compartment of sural and calcaneal regions. The experiment is then repeated for four other participants and the trends are observed. The experiment is extremely important as the primary knowledge gained will assists us to generate muscle-tendon units which can result into better understanding of the force and energy production.

    Paper
    Multi stage classification and segmentation of brain tumor
    3rd International Conference on Computing for Sustainable Global Development (INDIACom), 2016

    G.B. Praveen, Anita Agrawal

    A multi stage approach has been proposed to detect tumor, classify them into glioma or meningioma and perform their segmentation. In the first stage, preprocessing is performed, which includes noise filtering, image cropping, scaling operations and histogram equalisation. Feature extraction is performed in the second phase in which 19 features are extracted from Gray Level Co-occurrence Matrix, Gray Level Run Length Matrix and histogram based techniques which are the input to the classifier. Random forest classifier has been used to classify Magnetic Resonance Images into normal or abnormal and if abnormal, into glioma or meningioma. Final phase deals with the tumor detection performed using Fast Bounding Box and the tumor segmentation using active contour model.

    Paper
    Hybrid approach for brain tumor detection and classification in magnetic resonance images
    Communication, Control and Intelligent Systems (CCIS), 2015

    GB Praveen, Anita Agrawal

    A hybrid approach for brain tumor detection and classification through magnetic resonance images has been proposed. First phase of the proposed approach deals with image preprocessing which includes noise filtering, skull detection, etc. The second phase deals with feature extraction of MR brain images using gray level co-occurrence matrix. Third phase deals with classification of inputs into normal or abnormal using Least Squares Support Vector Machine classifier with Multilayer perceptron kernel. Final phase is the segmentation of the tumor part from the brain using fast bounding box.

    Paper
    [Reviewer]
    • Medical Image Computing and Computer Assisted Intervention (MICCAI), 2020, 2021, 2023
    • Machine Learning for Health (ML4H) workshop at Neural Information Processing Systems (NeurIPS), 2020, 2021, 2022
    • IEEE Transactions on Cybernetics
    • IEEE Journal of Biomedical and Health Informatics
    • IEEE Access
    • Computerized Medical Imaging and Graphics, Elsevier
    • Future Generation Computer Systems, Elsevier
    • Journal of Digital Imaging, Springer
    • Human-centric Computing & Information Sciences, Springer
    • Computers in Biology & Medicine, Elsevier
    • International Journal of Imaging Systems & Technology, Wiley
    • Informatics in Medicine Unlocked, Elsevier
    [Professional Activities]

    © Praveen G.B
    Last updated: March 2023