About Princeton Medical Systems

The Trust Layer for Medical Imaging AI Training Data

We author the VIDS standard. We operate as the commercial validator against it. Our work is to make medical imaging datasets defensible — in regulatory submissions, in procurement decisions, and in the real-world clinical use of the AI they train.

Our Story

From a Gap in the Market to an Open Standard

Princeton Medical Systems was founded in March 2024 in Chennai, India. We started where the medical AI industry's biggest bottleneck actually is: not models, but the data those models train on — inconsistent, poorly documented, often non-compliant with regulatory requirements. The cost of that gap is measured in delayed treatments, failed FDA submissions, and clinical AI deployments that fail under audit.

We built the missing layer in two parts. The first is VIDS — the Verified Imaging Dataset Standard — an open specification that defines what a defensible medical imaging dataset looks like. We authored it, published it under CC BY 4.0, and released the reference validator as open-source on PyPI.

The second is the commercial operating layer: validation services, reference datasets, and Compliance Evaluations that issue independent attestations buyers can put in front of regulators, procurement teams, and auditors. The standard is open. Our delivery against it is the commercial offering.

Our Mission

To make medical imaging datasets defensible — by building and operating the open standard that lets any buyer verify what they're acquiring, before they trust it.

Our Vision

Every clinical AI model is trained on data that has been independently validated against an open, auditable standard — and every buyer can verify that validation themselves.

Our Journey

Company Timeline

March 2024

Founded

Princeton Medical Systems established with a mission to solve medical imaging data quality.

Mid 2024

First DICOM Tools Launched

Released initial suite of DICOM conversion and de-identification tools for developers.

Late 2024

PMS Imaging Standard Development

Began authoring VIDS — the open Verified Imaging Dataset Standard — as the foundation of our commercial operating model. PMS Imaging Standard™ is our delivery against the open spec. Based on VIDS.

Early 2025

Talent Pipeline Established

Published Udemy courses and established structured training programs for medical imaging AI engineers.

January 2026

SJIT University Partnership

Signed MOU with SJIT University to advance medical imaging AI education and research.

2025

Dataset Curation Begins

Initiated radiologist-annotated dataset curation for lung CT and chest X-ray imaging.

April 2026

VIDS Public Release

Released VIDS v1.0 publicly: spec on vidsstandard.org under CC BY 4.0, reference validator on PyPI as open-source, LIDC-Hybrid-100 reference dataset on Zenodo, peer-reviewed preprint on arXiv. Launched the Compliance Evaluation service for medtech and AI buyers.

2026

Full Platform Expansion

Expanding the platform with new datasets, tools, and enterprise partnerships.

Leadership Team

Dr. Joan S. Muthu, Ph.D.

Co-Founder & Chief Technology Officer

Dr. Joan S. Muthu brings over 14 years of experience in DICOM engineering, medical image interoperability, and AI/ML for medical imaging. She holds a Ph.D. in Computer Science and Engineering, with doctoral research in DICOM file security that led to a granted Indian patent and a Gold Medal for Best Research Paper. Joan has published six peer-reviewed papers in SCI- and Scopus-indexed journals and serves as a peer reviewer for Springer Nature. As primary author of the VIDS specification, Joan led the open-source reference validator and the LIDC-Hybrid-100 reference dataset that demonstrate VIDS-Full compliance — establishing the technical foundation of PMS's commercial operating model.

At Princeton Medical Systems, Joan leads the development of end-to-end medical imaging data pipelines — from cross-format image conversion and interoperability to large-scale dataset curation and the company's proprietary dataset verification framework — ensuring every dataset is spatially accurate, metadata-complete, and fully traceable.

John Shalen R

Co-Founder & Chief Operating Officer

John Shalen R brings a distinctive combination of clinical data standards expertise, statistical programming, and hands-on operational leadership to the company's mission. He has direct experience in regulatory-aligned data workflows, having developed and validated clinical trial datasets to CDISC standards (SDTM & ADaM) using R and SAS. John previously built a company's team and operational infrastructure from the ground up as General Manager and Director at Orthogonus Technologies. At Princeton Medical Systems, John leads the commercial operating model for VIDS — from Compliance Evaluation services to medtech and AI buyer engagement — ensuring the standard's open foundation translates into defensible procurement workflows.

At Princeton Medical Systems, John leads proof-of-concept development, strategic planning, and institutional collaborations at the intersection of healthcare technology and clinical research.

Our Principles

What Drives Us

The Standard Is Open

VIDS is published under CC BY 4.0. The reference validator is open-source. Our reference dataset is freely available on Zenodo. Buyers can verify our methodology independently — before they trust our datasets, before they pay for an evaluation, before they put our work in front of regulators.

Independence Over Integration

We don't bundle validation with the datasets we sell. The same standard that governs our reference datasets is available to any aggregator, BPO, or in-house team that wants to validate their own data. Buyers pay for validation because it's independent — not because we made the data.

Defensibility Is the Product

A dataset that works in development but can't be defended in an FDA submission, a procurement audit, or a vendor dispute is not a finished product. We design every Validation Report and Validation Attestation to hold up under scrutiny — because that's where they're going to be used.

Join Us

We're building the trust layer for an industry where trust is the binding constraint. If you care about medical imaging, AI training data, or building the standards that make AI defensible — we'd like to talk.

careers@princetonmedicalsystems.com