Arun Innanje
Senior Principal Software Architect & Head of AI Applications ·
UII America Inc.,
Cambridge, MA
I lead AI research, product architecture, and clinical deployment at United Imaging Intelligence (UII) — the AI subsidiary of United Imaging Healthcare (SSE: 688271), a publicly traded global leader in medical imaging and radiotherapy equipment with $1.7B+ in annual revenue across 85+ countries. UII develops FDA-cleared and CE-marked AI-powered medical devices deployed in hospitals worldwide.
My work spans privacy-preserving federated learning, cardiac MRI analysis (validated with Washington University Medical School), AI-powered patient positioning systems, and clinical AI deployments across CT, MRI, X-ray, and DR modalities. Previously, I spent 14 years at Siemens Corporation (Princeton, NJ), growing from software engineer to architect and technical lead.
B.E. in Computer Science, RNS Institute of Technology, Visvesvaraya Technological University, India.
50+
Patent Filings
25
Granted US Patents
595+
Citations
20+
Years Experience
News
2026
Paper accepted at CVPR 2026: "MedGRPO: Multi-Task Reinforcement Learning for Heterogeneous Medical Video Understanding".
Dec 2025
Dec 2025
US Patent granted: "Apparatus and Method to Enable Low Power MR Patient Positioning on Edge Devices" (US12487305B2).
Nov 2025
RSNA 2025 abstract accepted (with Mass General Brigham): "Validation of an AI Tool in Detection, Localization, and Characterization of Osseous Metastases on Thoracic CT Examinations."
2025
SCMR 2025 abstract published in Journal of Cardiovascular Magnetic Resonance (with Washington University Medical School): Fully automated cardiac function & strain analysis achieving ICC up to 99% with expert observers.
2025
Serving on Program Committee of APAH 2025 Workshop at ICCV 2025, Honolulu, Hawaii.
2026
Serving on Program Committee of AIBSD 2026 Workshop at AAAI 2026, Singapore.
Jul 2025
US Patent granted: "Patient Care Record Management System" (US12361161B2) — 8th patent as lead inventor.
Feb 2025
US Patent granted: "Anatomy-Aware Contour Editing Method and System" (US12229954B2).
Research Interests
My research explores the intersection of artificial intelligence and medical imaging, with a focus on building clinically deployable, privacy-preserving AI systems that perform reliably across institutions and modalities.
Publications
Journal Articles
1
Federated Learning With Privacy-Preserving Ensemble Attention Distillation
IEEE Transactions on Medical Imaging, Vol. 42, pp. 2057–2067, 2022
3
An Evaluation of Zero-Click Whole-Workflow of Cardiovascular Magnetic Resonance Cardiac Function and Strain Analyses Pipeline on Cardio-Oncology Patients
EMJ Radiology (European Medical Group)
4
AI-Assisted Soft Tissue Virtual Planning in Fibula Free Flap Jaw Reconstruction
Plastic and Reconstructive Surgery (Lippincott Williams & Wilkins)
Conference Proceedings (Peer-Reviewed)
5
MedGRPO: Multi-Task Reinforcement Learning for Heterogeneous Medical Video Understanding
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2026)
6
Ensemble Attention Distillation for Privacy-Preserving Federated Learning
IEEE/CVF International Conference on Computer Vision (ICCV), pp. 15056–15066, 2021
RSNA Annual Meeting Scientific Abstracts
9
Clinical and Imaging Factors Associated with Growth of Subsolid Pulmonary Nodule on CT
RSNA 2023 Annual Meeting, Chicago, IL
10
Clinical & Imaging Predictor of Subsolid Nodule Growth
RSNA 2023 Annual Meeting, Chicago, IL
11
Radiomics Analysis for Predicting Progression of Part-Solid Nodules on CT
RSNA 2023 Annual Meeting, Chicago, IL
12
Validation of an AI Tool in Detection, Localization, and Characterization of Osseous Metastases on Thoracic CT Examinations
RSNA 2025 Annual Meeting, Chicago, IL
SCMR Scientific Session Abstract
13
Comprehensive Deep Learning-Automated Analysis of Cardiac Function and Myocardial Strain Using CMR in Clinical Setting: Achieving Comparable Accuracy to Expert Observers
Journal of Cardiovascular Magnetic Resonance, Vol. 27, 101357, 2025
Granted U.S. Patents
25 granted · 8 as first/lead inventor · 17 as co-inventor · 25+ additional applications pending · 1 European patent (EP-3655923-B1)
First / Lead Inventor (8 Patents)
| # | Title | Patent No. | Date |
|---|---|---|---|
| 1 | Patient Care Record Management System | US12361161B2 | Jul 2025 |
| 2 | Anatomy-Aware Contour Editing Method and System | US12229954B2 | Feb 2025 |
| 3 | Click Based Contour Editing | US12141420B2 | Nov 2024 |
| 4 | Orientation Detection in Overhead Line Insulators | US11861480B2 | Jan 2024 |
| 5 | Plugin and Dynamic Image Modality Reconstruction Interface Device | US11756240B2 | Sep 2023 |
| 6 | Systems and Methods for Managing Medical Data Using Relationship Building | US11734333B2 | Aug 2023 |
| 7 | Cardiac Feature Tracking | US11315246B2 | Apr 2022 |
| 8 | Methods and Devices for Enhancing Functionalities Related to Sensors | US10887558B1 | Jan 2021 |
Co-Inventor (17 Patents)
| # | Title | Patent No. | Date |
|---|---|---|---|
| 9 | Apparatus and Method to Enable Low Power MR Patient Positioning on Edge Devices | US12487305B2 | Dec 2025 |
| 10 | Systems and Methods for Annotating 3D Data | US12444171B2 | Oct 2025 |
| 11 | Systems and Methods for Fetus Monitoring | US12178641B2 | Dec 2024 |
| 12 | Sensing Device for Medical Facilities (continuation) | US12138015B2 | Nov 2024 |
| 13 | System and Methods for Privacy Preserving Cross-Site Federated Learning | US12045695B2 | Jul 2024 |
| 14 | Key Points Detection Using Multiple Image Modalities | US11967102B2 | Apr 2024 |
| 15 | Methods and Devices for Guiding a Patient (continuation) | US11836946B2 | Dec 2023 |
| 16 | System and Method for Configuring Internet Protocol Device | US11811729B1 | Nov 2023 |
| 17 | Systems and Methods for Image Style Transformation | US11625576B2 | Apr 2023 |
| 18 | Systems and Methods for Machine Learning Based Modeling | US11604984B2 | Mar 2023 |
| 19 | Abnormality Detection Within a Defined Area | US11386537B2 | Jul 2022 |
| 20 | Systems and Methods for Enhancing a Distributed Medical Network | US11379727B2 | Jul 2022 |
| 21 | Sensing Device for Medical Facilities | US11335456B2 | May 2022 |
| 22 | Methods and Devices for Guiding a Patient | US11257251B2 | Feb 2022 |
| 23 | Systems and Methods for Imaging | US11244446B2 | Feb 2022 |
| 24 | Weakly Supervised Anomaly Detection and Segmentation in Images | US10672115B2 | Jun 2020 |
| 25 | Part Identification Using a Photograph and Engineering Data | US10073848B2 | Sep 2018 |
European Patent also granted: EP-3655923-B1
Experience
Senior Principal Software Architect — Head of AI Applications & Infrastructure
Mar 2019 – Present
UII America Inc. (United Imaging Intelligence) · Cambridge, MA
- Technical lead for multimillion-dollar, multiyear projects, directing 10+ engineers across the US, Germany, and India
- Principal architect of federated learning infrastructure enabling privacy-preserving multi-institutional collaboration with Mass General Brigham and UCSF
- Architected Cardiac MRI Function & Feature Tracking application (Vue.js + microservices) for comprehensive cardiac function analysis and myocardial strain calculations from cine MRI data — validated with Washington University Medical School for cardio-oncology applications
- Designed uAI Vision (FDA/CE pathway) — camera-based AI system for automated patient positioning, integrated into CT, X-ray, DR, and MR product pipelines deployed in hospitals worldwide
- Built and manages semi-HPC GPU cluster (V100, A100, H100) with Kubernetes orchestration
- Lead architect of Smart Annotation Tool (SAT) for AI-powered multi-modality medical image annotation
- Lead architect of MERITS (Metaverse Ecosystem for Robotics, Imaging, Therapeutics and Surgery) — mixed-reality surgical planning application for preoperative vessel and bone identification
- Drives research collaborations with top-tier academic medical centers resulting in peer-reviewed publications and RSNA presentations
Software Engineer → Software Architect → Technical Lead
May 2011 – Feb 2019
Siemens Corporate Research → Siemens Corporation · Princeton, NJ
- Industrial Imaging Plugin (IIP) — Technical Lead ($3M budget). Microservices platform for 2D/3D visual sensor data analytics with browser-based visualization and experiment orchestration (Python, Flask, Angular, Docker, AWS)
- SAMI — Software Architect ($1.2M). Enterprise NDT image management application for Siemens Energy service engineering (C++, Python, Qt, VTK)
- MR Compressed Sensing / Sparse Reconstruction — Converted research algorithms to production C++ with GPU acceleration (OpenMP, Intel MKL). Recipient of the R&D 100 Special Recognition Gold Award (2017)
- Easy Spares iDEA — Full-stack developer ($2M). AI-powered spare parts identification for Siemens digital supply chain using computer vision — identification in under 10 seconds (C++, C#, Python, OpenCV, Caffe)
- Forensic Video Analytics — Multi-camera crowd analysis and individual re-identification for intelligent surveillance (C++, Python, HTML5)
- MR Neuro3D (syngo.via) — Functional neuroimaging including BOLD fMRI and Diffusion Tensor Imaging (DTI); registration, brain mask, VOI, and functional analysis plots
Software Engineer → Architect
Jun 2005 – May 2011
Siemens Technology Solutions · Bangalore, India
- MR Spectroscopy (syngo.via) — Architect. Next-generation syngo.via-based spectroscopic visualization of MR scan results; integrated as standalone taskflow and into existing MR workflows
- Enterprise CAD Solutions (ECS) — Software Engineer. syngo.via-based application for multi-timepoint processing and viewing of patient scans with CAD result visualization
- TrueD (syngo classic) — Software Developer (3 years). Siemens-only solution for temporal comparison of PET/CT and SPECT/CT studies across timepoints for therapy assessment. Developed modules for visualization, registration, layout editing, gated support, VOI, and multi-timepoint quantification
Awards & Distinctions
R&D 100 Special Recognition Gold Award — 2017
Awarded for MR Compressed Sensing Sparse Reconstruction technology at Siemens Healthineers. A breakthrough enabling significantly faster MRI scan times with preserved image quality. The R&D 100 Awards are among the most prestigious international honors in applied research and technology development.
Certified PM@CT (Project Manager) — Siemens AG
One of a select number of engineers globally chosen for cross-functional project management certification by Siemens AG.
Professional Service & Affiliations
- Program Committee Member — AIBSD 2026: Workshop on Artificial Intelligence with Biased or Scarce Data, at the 40th AAAI Conference on Artificial Intelligence, Singapore
- Program Committee Member — APAH 2025: Workshop on Advanced Perception for Autonomous Healthcare, at IEEE/CVF ICCV 2025, Honolulu, Hawaii
- Peer Reviewer — Evaluating scientific submissions on AI under data bias and scarcity, federated learning, trustworthy AI, and healthcare AI applications
- Active contributor to RSNA annual meetings through scientific abstract submissions and industry presentations (~50,000 attendees, world's largest radiology conference)
Technical Expertise
- AI / ML
- PyTorch, TensorFlow, Federated Learning, Deep Learning, Computer Vision, Radiomics, Medical Image Segmentation, LLMs, RAG, Reinforcement Learning
- Languages
- Python, C++, C#, JavaScript, HTML5, CSS
- Frameworks
- Vue.js, Angular, Flask, .NET, Docker, Kubernetes, Ansible
- Cloud & Infra
- AWS, GPU Cluster Management (Slurm, Open OnDemand), Cloudflare
- Medical Imaging
- DICOM, VTK, OpenCV, syngo.via, Medical Device (FDA/CE) Workflows
- Databases
- PostgreSQL, Oracle, CouchDB, Elasticsearch
- Project Mgmt
- Certified PM@CT (Siemens), Agile, Portfolio Management