FlowCV Logo
Adway KanhereMedical Image Analysis | Artificial Intelligence
Education
The Johns Hopkins University, USA, Master of Science in Engineering Degree, Biomedical Engineering
2021- 2023

Specialization in Biomedical Data Science with Thesis

M.S Ramaiah Institute of Technology, India, Bachelor of Engineering Degree, Medical Electronics
2016 - 2020
Work Experience
University of Maryland School of Medicine, USA & University of Maryland Institute of Health Computing, USA (Jul 2024-), Bioinformatics Software Engineer-II
Jun 2023 – present | Full-Time
  • •Purpose: Lead engineering projects to build automated AI solutions that enhance clinical workflows and improve patient care.

  • •Tech Stack: PyTorch, FastChat, Docker, Apache Airflow, Kubernetes

  • •Responsibilities and Outcomes:
  • •Setup and deployed a medical imaging provisioning toolkit based on the Kaapana backend using a custom container registry.
  • •Investigated the use of LLaVa-Med and GPT4-Vision for visual question answering of diagnostic conditions using radiological images.
  • •Developed and deployed natural language processing pipelines leveraging Vicuna 7B and 13B models backed by very large language models (VLLMs) to extract Agatston coronary artery calcium scores from radiology reports using prompt engineering.
  • •Configured and deployed an online DICOM Viewer “Vision” for visualizing 3D DICOM scans based on OHIF, setup connectivity to custom DICOMWeb server backend. Added automated reporting, segmentation, and active learning capabilities using MONAILabel.
  • •Contributed to developing an existing codebase to pull imaging data from PACS/VNA, automate anonymization of image data in accordance to HIPAA.
  • •Developed “Maverick”, an AutoML tool for medical image analysis and segmentation - integrated with Weights&Biases and built using Streamlit.
    • University of Maryland School of Medicine, USA, Bioinformatics Software Engineer - I
    Sep 2022 – May 2023 | Part-Time
  • •Purpose:
  • •Defined, maintained, and automated software pipelines for medical image data workflows from ingestion to model development for clinically translatable segmentation tasks.
  • •Assisted in authoring technical papers and patents to communicate key innovations.

  • •Tech Stack: Python, PyTorch, AWS, GCP, Azure OpenAI, LlamaIndex

  • •Responsibilities and Outcomes:
  • •Developed and implemented federated learning algorithm “SegViz” for collaborative segmentation from heterogeneous medical image datasets with incomplete annotations. Led partnership with industry experts from Flower labs to push to production.
  • •Ideated and assisted in developing "Text2Cohort", an LLM toolkit that allows users to interact with the imaging data commons (IDC) using natural language.
  • •Developed code using LlamaIndex, LangChain, and OpenAI API to load structured data and create a vector database for Retrieval Augmented Generation on IDC metadata.
  • •Developed code pipelines in medical image segmentation for clinical applications to several collaborators within the UM Medical system.
  • •Standardized data and AI infrastructure pipelines with cloud and on-prem resources to enable high throughput scalable computing.
    • Regeneron Pharmaceuticals, USA, Data Science Intern - Early Clinical Development
    Jun 2022 – Aug 2022 | Internship
  • •Purpose: Apply data science and deep learning to advance oncology research and improve drug response.

  • •Tech Stack: Python, PyTorch, nnUNet, Streamlit

  • •Responsibilities and Outcomes:
  • •Developed a deep learning based pipeline for automated segmentation of lung tumor volumes from lung CT images on the NSCLC-Radiomics (TCIA) dataset.
  • •Developed a CNN based regression pipeline for progression-free survival, overall survival prediction and response to immunotherapy for patients with non-small cell lung cancer.
  • •Designed and deployed an internal web application based GUI to translate the trained AI model into a tool that can be used interactively by scientists.
    • Johns Hopkins University & School of Medicine, Graduate Research Assistant
    Dec 2021 – May 2022 | Part-Time
  • •Purpose: Conduct deep learning research leveraging generative AI to advance medical imaging capabilities with smartphone devices.

  • •Tech Stack: Python, PyTorch, Detectron2, TorchXRayVision

  • •Responsibilities and Outcomes:
  • •Programmed a deep learning-based pipeline for transforming smartphone-based photos of chest X-Ray images to digital X-Ray images using a Pix2Pix Generative Adversarial Network (GAN).
  • •Programmed a deep learning-based pipeline for localizing the region of the thorax in chest X-Ray images using Detectron2's pipeline.
  • •Analyzed the performance of existing state-of-art CNN-based algorithms in TorchXRayVision to generalize the prediction of lung abnormalities on smartphone photographs of chest X-rays compared to synthetically modified digital chest X-Ray images.
    • Symbiosis Centre for Medical Image Analysis, India, Computational Neuroimaging Intern
    Jun 2019 – Aug 2019 | Internship
  • •Tech Stack: Pytorch, FSL, FreeSurfer, TorchIO
  • •Outcomes:
  • •Investigate deep learning capabilities at predicting regions of focal cortical dysplasia for MR+/MR- cases.
  • •Developed a CNN predictive model for locating regions of epilepsy on multimodal Magnetic Resonance images using AI and deep learning.
  • •Constructed an automated workflow for skull stripping, co-registration, cortical surface, and volume map reconstruction for high-dimensional neuroimaging nuclear medicine data
    • MDS Bio-Analytics, India, Data Analytics Research Intern
    Jul 2018 – Aug 2018 | Intership
  • •Tech Stack: R/R Studio
  • •Outcomes: Evaluated hospital EHR data on tuberculosis by applying statistical modeling algorithms to model patient health data, and further test for outliers.
    • Peer-reviewed publications
    * First/Co-first author

    Kanhere, A.*, Kulkarni, P., Yi, P., & Parekh, V. (2024). Privacy-Preserving Collaboration for Multi-Organ Segmentation via Federated Learning from Sites with Partial Labels. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 2380–2387). (Best Poster Award)

    Kanhere, A.*, Navarathna, N.*, Yi, P., Parekh, V., Pickle, J., Luber, J., Cloak, C., Ernst, T., Chang, L., Li, D., Redline, S., & Isaiah, A. (2024). Unlocking Hidden Data: Deep Learning for Opportunistic Airway Analysis in Large-Scale Pediatric Neuroimaging. American Journal of Respiratory and Critical Care Medicine. (Under Review)

    Chatterjee, D.*, Kanhere, A.*, Doo, F., Zhao, J., Chan, A., Welsh, A., Kulkarni, P., Trang, A., Parekh, V., & Yi, P. (2024). Children Are Not Small Adults: Addressing Limited Generalizability of an Adult Deep Learning CT Organ Segmentation Model to the Pediatric Population. Journal of Imaging Informatics in Medicine, 1–14.

    Kamel, P.*, Khalid, M., Steger, R., Kanhere, A., Kulkarni, P., Parekh, V., Yi, P., Gandhi, D., & Bodanapally, U. (2024). Dual Energy CT for Deep Learning-Based Segmentation and Volumetric Estimation of Early Ischemic Infarcts. Journal of Imaging Informatics in Medicine, 1–12.

    Doo, F.*, Savani, D.*, Kanhere, A., Carlos, R., Joshi, A., Yi, P., & Parekh, V. (2024). Optimal Large Language Model Characteristics to Balance Accuracy and Energy Use for Sustainable Medical Applications. Radiology, 312(2), e240320.

    Kargilis, D.*, Kulkarni, P., Kanhere, A., Garin, S., Murphy, Z., Hafey, C., Parekh, V. S., & Yi, P. H. (2024). The Impact of Standard Image Preprocessing on Deep Learning Models’ Predictions for Chest Radiographs: An Overlooked Source of Performance Variability. Journal of Imaging Informatics in Medicine (Under Review).

    Kamel, P.*, Kanhere, A., Kulkarni, P., Khalid, M., Steger, R., Bodanapally, U., Gandhi, D., Parekh, V., & Yi, P. (2024). Optimizing Acute Stroke Segmentation on MRI Using Deep Learning: Self-Configuring Neural Networks Provide High Performance Using Only DWI Sequences. Journal of Imaging Informatics in Medicine, 1–10.

    Kulkarni, P.*, Kanhere, A., Siegel, E., Yi, P., & Parekh, V. (2024). ISLE: An Intelligent Streaming Framework for High-Throughput AI Inference in Medical Imaging. Journal of Imaging Informatics in Medicine, 1–14.

    Kulkarni, P.*, Kanhere, A., Yi, P. H., & Parekh, V. S. From Isolation to Collaboration: Federated Class-Heterogeneous Learning for Chest X-Ray Classification. Proceedings of the 4th Machine Learning for Health Symposium (ML4H), 2024.

    Kulkarni, P.*, Kanhere, A., Kukreja, H., Zhang, V., Yi, P. H., & Parekh, V. S.# (2024). Improving Multi-Center Generalizability of GAN-Based Fat Suppression using Federated Learning. MIDL 2024 (Medical Imaging with Deep Learning): Short Paper Track.

    Kulkarni, P.*, Kanhere, A.*, Savani, D.*, Chan, A., Chatterjee, D., Yi, P. H., & Parekh, V. S.# (2024). Anytime, Anywhere, Anyone: Investigating the Feasibility of Segment Anything Model for Crowd-Sourcing Medical Image Annotations. MIDL 2024 (Medical Imaging with Deep Learning): Short Paper Track

    Yi, P. H.*, Bachina, P., Bharti, B., Garin, S. P., Kanhere, A., Kulkarni, P., Li, D., Parekh, V. S., Santomartino, S. M., Moy, L., & Sulam, J. (2024). Pitfalls and Best Practices in Evaluation of Algorithmic Biases in Radiology. Radiology (In Press).

    Doo, F.*, Parekh, V., Kanhere, A., Savani, D., Tejani, A., Sapkota, A., & Paul, H. (2024). Evaluation of climate-aware metrics tools for radiology informatics and artificial intelligence: toward a potential radiology ecolabel. Journal of the American College of Radiology, 21(2), 239–247.

    Bachina, P.*, Garin, S., Kulkarni, P., Kanhere, A., Sulam, J., Parekh, V., & Yi, P. (2023). Coarse race and ethnicity labels mask granular underdiagnosis disparities in deep learning models for chest radiograph diagnosis. Radiology, 309(2), e231693.

    Kulkarni, P.*, Kanhere, A., Yi, P. H., & Parekh, V. S. (2022). From Competition to Collaboration: Making Toy Datasets on Kaggle Clinically Useful for Chest X-Ray Diagnosis Using Federated Learning. Medical Imaging meets NeurIPS Workshop, 2022 Conference on Neural Information Processing Systems

    PrePrints

    Kulkarni, P.*, Kanhere, A., Siegel, E., Yi, P. H., & Parekh, V. S. (2023). One Copy Is All You Need: Resource-Efficient Streaming of Medical Imaging Data at Scale. arXiv preprint arXiv:2307.00438.

    Kulkarni, P.*, Kanhere, A., Yi, P. H., & Parekh, V. S. Text2Cohort: Facilitating Intuitive Access to Biomedical Data with Natural Language Cohort Discovery. arXiv preprint arXiv:2305.07637.

    CONFERENCE PRESENTATIONS

    * Presenting author

    Kanhere, A.*, Kulkarni, P., Yi, P., & Parekh, V. (2024). Privacy-Preserving Collaboration for Multi-Organ Segmentation via Federated Learning from Sites with Partial Labels. Data Curation and Augmentation in Medical Imaging Workshop at CVPR 2024. (Best Poster Award)

    Kamel, P.*, Khalid, M., Steger, R., Kanhere, A., Kulkarni, P., Parekh, V. S., Yi, P. H, Bodanapally, U., & Gandhi, D. (2024). Is Dual-Energy CT Better for Deep Learning-Based Detection and Segmentation of Early Ischemic Infarcts on CT? 2024 American Society of Neuroradiology Annual Meeting.

    Kamel, P.*, Kanhere, A., Kulkarni, P., Khalid, M., Steger, R., Bodanapally, U., Gandhi, D., Parekh, V. S., & Yi, P. H. (2024). Assessing the Generalizability of Acute Stroke Segmentation using a Self-Configuring Neural Network Trained on Public Data. 2024 American Society of Neuroradiology Annual Meeting.

    Kamel, P.*, Khalid, M., Steger, R., Kanhere, A., Kulkarni, P., Parekh, V. S., Yi, P. H., Bodanapally, U., & Gandhi, D. (2024). Cross-Modality Stroke Segmentation using Deep Convolutional Neural Networks for Detection of Acute Ischemic Infarcts on Non-Contrast Head CT. 2024 American Society of Neuroradiology Annual Meeting.

    Navarathna, N.*, Chatterjee, D., Chan, A., Kulkarni, P., Kanhere, A., Parekh, V. S., & Yi, P. H. (2024). From Download to ML: Challenges in Directly Using the MIDRC Dataset for Machine Learning and Enhancing its Usability. Spotlight Talk, 2024 Society for Imaging Informatics in Medicine Annual Meeting.

    Chatterjee, D., Kanhere, A., Trang, A., Parekh, V., Yi, P. (2023). Augmenting the MIDRC Dataset using Deep Learning-Based Quantification of Abdominal Aortic Calcification: Proof-of-Concept for Population-Level Disease Screening, Podium presentation, 2023 Conference on Machine Intelligence in Medical Imaging, Society for Imaging Informatics in Medicine, Baltimore, MD

    Chatterjee, D., Kanhere, A., Shoum, B L., Singh, S, Parekh, V.S, Yi, P.H., Arbab-Zadeh, A. (2023). Automated Detection of Pericoronary Adipose Tissue Attenuation to Detect Inflammation on Coronary Computed Tomography Angiography, Podium presentation, 2023 Conference on Machine Intelligence in Medical Imaging, Society for Imaging Informatics in Medicine, Baltimore, MD

    Zhao, J.*, Kanhere, A., Kulkarni, P., Chatterjee, D., Parekh, V. S., & Yi, P. H. (2024). Using Deep Learning to Predict Knee Osteoarthritis. Poster Presentation, Undergraduate Research Day 2024, University of Maryland.

    Kim, J.*, Kulkarni, P., Welsh, A., Garin, S., Chatterjee, D., Kanhere, A., Parekh, V. S, & Yi, P. H. (2023). Sex Bias in Pediatric Deep Learning Chest Radiograph Classifier Model. Spotlight Talk, Medical Student Research Day 2023, University of Maryland-Baltimore.

    Kamel, P.*, Kanhere, A., Kulkarni, P., Khalid, M., Steger, R., Bodanapally, U., Gandhi, D., Parekh, V. S., & Yi, P. H. (2023). Quantifying the Technical Challenges and DICOM Metadata Variability in Stroke Machine Learning Data Curation. Spotlight Talk, Radiological Society of North America 109th Scientific Assembly and Annual Meeting.

    Kulkarni, P.*, Kanhere, A., Yi, P. H., & Parekh, V. S. (2023). From Isolation to Collaboration: Harmonizing Heterogeneous Medical Imaging Datasets with Partial Annotations. Spotlight Talk, 2023 Conference on Machine Intelligence in Medical Imaging, Society for Imaging Informatics in Medicine.

    Bachina, P.*, Garin, S., Kulkarni, P., Kanhere, A., Sulam, J., Parekh, V. S., & Yi, P. H. (2023). Coarse Race and Ethnicity Labels Mask Granular Underdiagnosis Disparities in Deep Learning Models for Chest Radiograph Diagnosis. Spotlight Talk, 2023 Conference on Machine Intelligence in Medical Imaging, Society for Imaging Informatics in Medicine

    Kulkarni, P.*, Kanhere, A., Siegel, E., Yi, P. H., & Parekh, V. S. (2023). One Copy Is All You Need: Resource-Efficient Streaming of Medical Imaging Data at Scale. Poster Presentation, 2023 Conference on Machine Intelligence in Medical Imaging, Society for Imaging Informatics in Medicine.

    Kulkarni, P.*, Kanhere, A., Yi, P. H., & Parekh, V. S. (2023). Text2Cohort: Democratizing the NCI Imaging Data Commons with Natural Language Cohort Discovery. Poster Presentation, 2023 Conference on Machine Intelligence in Medical Imaging, Society for Imaging Informatics in Medicine.

    Kanhere, A.*, Kulkarni, P., Yi, P. H., & Parekh, V. S. (2023). SegViz: A Federated Learning Framework to Train Multi-task Segmentation Models from Partially Annotated and Distributed Datasets. Poster Presentation, 2023 Conference on Machine Intelligence in Medical Imaging, Society for Imaging Informatics in Medicine.

    Bachina, P.*, Garin, S., Kulkarni, P., Kanhere, A., Kargilis, D., Parekh, V. S., & Yi, P. H. (2023). Not So Black and White: Confounders Mediate AI Prediction of Race on Chest X-Rays. Poster Presentation, Machine Learning for Healthcare 2023.

    Kamel, P.*, Kanhere, A., Kulkarni, P., Parekh, V. S., & Yi, P. H. (2023). Optimizing Acute Stroke Segmentation: Do Additional Sequences Matter for Deep Learning Algorithms? Poster Presentation, 2023 Society for Imaging Informatics in Medicine Annual Meeting.

    Patents

    Parekh, V. S., Kulkarni, P., Kanhere, A., Yi, P. H., & Siegel, E. Systems and Methods for High-Throughput Analysis for Graphical Data. US Patent Application No. 63/501,552 – Filed May 11, 2023. PCT International Patent Application No. PCT/WO2024/233969 - Filed November 14, 2024.

    Book Chapters

    Parekh, V. S., Kulkarni, P., Kanhere, A., & Jacobs, M. A. (2024). Expanding the Federated Horizon: Cross-Domain Techniques for Collective Intelligence. Federated Learning for Medical Imaging: Principles, Algorithms and Applications, The MICCAI Society Book Series (In Press).

    Invited Talks
    Privacy preserving 3D medical image segmentation using Flower, Flower Monthly Series 2024-07, Flower Labs, UK
    Teaching experience
    Johns Hopkins University - Carey School of Business, USA, Adjunct Instructor
    Feb 2024 – present
  • •Purpose: Teach machine learning to MS in Business Analytics and Risk Management and Financial Engineering students
  • •Responsibilities: Big Data Machine Learning - Spring 2024 (1 section); Practical Machine Learning - Spring 2025 (2 sections)
    • Johns Hopkins University - Carey School of Business, USA, Graduate Teaching Assistant
    Mar 2022 – May 2023
  • •Purpose: Support MS business management student learning and enhancement of machine learning skills.
  • •Responsibilities:
  • •Supported student learning in over 36 machine learning hackathons, enriching the educational experience for 150+ MS business analytics and risk management students.
  • •Sourced relevant datasets, developed interactive Jupyter notebooks with starter code, and crafted engaging machine learning challenges to enable hands-on practice.
  • •Received outstanding student feedback for patient and effective teaching style, professional demeanor, and commitment to student mastery of machine learning concepts and tools.
  • •Courses assisted: Big Data, Machine Learning (Spring II 2022); Operations Management (Fall I 2022); Data Analytics in R (Fall II 2022); Linear Econometrics for Finance (Fall II 2022); Python for Data Analytics (Spring I 2023); Empirical Finance (Spring II 2023); Big Data, Machine Learning (Spring II 2023) - Head TA
    • Awards
    Best Poster Award, Data Curation and Augmentation in Medical Imaging (DCAMI) Workshop at CVPR 2024
    Ranked 1st ($5000 Prize) HOPSTART 2022, Johns Hopkins New Venture Challenge
    Ranked 1st ($3000 Prize) FastForwardU Spark Incubator Showcase 2022, Johns Hopkins Technology Ventures (JHTV)
    National Talent Search Exam, State of Karnataka, India 2014

    Ranked #13 of 179 finalists & 80,000 participants

    Certificates
    Mentoring
    Nithya Navarathna, Research Co-ordinator, University of Maryland School of Medicine 2023-Present: Research mentor and lead for her work on data curation using the MIDRC dataset, linking upper airways with cognitive outcomes, and her work on functional near-infrared spectroscopy resulting in co-first author submission to AJRCCM and a conference presentation at SIIM
    Dharmam Savani, MS CS Student, University of Florida 2023-2023: Research mentor and lead for his work on open-source LLMs, building LLM infrastructure, cloud ecological implication in healthcare resulting in a publication in Radiology.
    Issac Chaudry, MD-PhD Student, University of Maryland Baltimore 2023-2023: Research mentor for his work on radiomics informed connectomics analysis on medical image segmentations.
    Vivian Zhang, MD Student, University of Maryland Baltimore. 2022-2023: Research mentor for her work on using GANs for converting fat suppressed to non-fat suppressed MRIs, classifying glioblastomas on brain MRIs and gross tumor resection resulting in 2 conference presentations at MIDL
    Devina Chatterjee, MD Student, University of Maryland Baltimore. 2023-2023: Research, technical mentor, and project lead for her work on applying AI to pediatric imaging and detection of pericoronary fat attenuation on CTCA, resulting in one journal publication at JIIM and several conference presentations at CMIMI.
    Alexander Welsh, MD Student, University of Maryland Baltimore. 2023-2023: Research mentor for his work on using saliency maps for AI interpretability of chest radiographs.
    Jake Kim, MD Student, University of Maryland Baltimore. 2023-2023: Research mentor for his work on bias in DL models of chest radiographs
    Andrew Chan, BS Student, University of Maryland, College Park 2023-2023: Research mentor for his work on AI fairness presented at SIIM and on transfer learning for pediatric imaging
    Jerry Zhao, BS Student, University of Maryland, College Park 2023-2023: Research mentor for his work on using deep learning to predict knee osteoarthritis using nnUNet and OAR dataset
    Noam Fox, BS Student, University of Maryland, College Park 2023-2023: Research mentor for her work on using nnUNets for segmenting the knee bone and cartilage on MR images.
    Preetham Bachina, MD Student, Johns Hopkins School of Medicine 2022-2023: Research mentor on his work on race and ethnicity labels for health disparities on chest radiographs which resulted in several conference abstracts and one publication in Radiology
    Dan Kargilis, MD Student, Johns Hopkins School of Medicine 2022-2023: Research mentor on his fairness in AI projects and assisted him in using our lab's cloud infrastructure.
    Sam Santomartino, Drexel University 2022-2024: Research mentor on her work on bone age prediction, interpretability of saliency maps on chest radiographs, and her coding skills.
    Open source contributions
    Building a Production-Level Federated Learning Framework for 3D Medical Image Segmentation, Flower Next Pilot Program, Flower Labs UK
    Jul 2023 – present
  • •Developed a Federated Learning system using the Flower API to create an FL setup for biomedical image segmentation using 3D-UNets using the MONAI library and data from the Medical Segmentation Decathlon Challenge.
  • •Wrote code, created documentation, and ran experiments in collaboration with computer scientists from Flower Labs
  • •Deployed a production environment for a demo at CVPR 2024
    • When are Deep Networks really better than Decision Forests at small sample sizes, and how?, Neuro Data Lab - Johns Hopkins Biomedical Engineering
    Aug 2021 – May 2022
  • •Assessed the conceptual & empirical comparisons between decision forests & deep networks for audio data on the FSDK-18 dataset.
  • •Remodeled the existing codebase to standardize and speed up the loading and pre-processing of audio data.
  • •Created a pipeline for Bayesian Hyperparameter tuning of the existing CNN models and improved the performance of the baseline CNN from 64% to 88% accuracy. Extended the same pipeline to improve the performance of all CNN models implemented in the previous study.
    • Volunteer Experience
    Gradvine, Graduate Mentor for Foreign Studies
    Aug 2021 – present
  • •Mentored students from engineering schools in India about applying to graduate engineering programs in biomedical engineering and biotechnology in the USA.
  • •Assisted in reviewing admissions essays for students.
  • •Advised students to plan their educational expenses and program of study.