Pixelence — removing a barrier most people don’t know exists.

How it started

The idea came from a hospital room. Less than 24 hours after my heart stopped, my LinkedIn photo was taken. A second cardiac event a few years later reminded me what the data looks like from the inside. I’d watched how much cost, how much risk, and how much time patients absorb for a diagnostic decision that should be simple.

When I joined Antler Malaysia’s MY2 cohort in March 2024, I told my coach in the first week that I wanted to build something in preventive healthcare. My first idea was rejected repeatedly — a symptom-logging app for early warning signs. Antler’s feedback was sharp and correct: behavior-dependent, hard to scale, hard to monetize. I had to let it go.

By week two I was working with Raheel Zubairi, who became my co-founder. We iterated through several more directions, most of them too broad. The turning point was narrowing radiology down to a single, honest question: which specific scan causes the most friction today, and can we remove that friction without making the diagnosis worse?

That question led us to contrast-enhanced brain MRI, and to Amr Mohammad — a clinical oncologist with published research in exactly this domain. I asked Antler to allow him to join as our third co-founder despite being external to the program. They agreed. We made it to the investment finals.

We didn’t get selected for investment. The committee’s comment was that the market wasn’t big enough — and that’s a message worth sitting with honestly. Either they’re right, or we didn’t deliver the message clearly enough. We decided we hadn’t. Eighteen months later, the data supports that call.

The barrier

Gadolinium-based contrast is injected into patients before a brain MRI so that tumors and lesions become visible on the scan. It works. It’s also a rare-earth metal that requires a functioning kidney, costs substantially more than the base scan, and leaves residual deposits in brain tissue that recent research is only beginning to characterize.

The patient journey looks like this: a baseline MRI, then a kidney function test, then a prescription for the Gadolinium injection, then a wait, then a second scan. About two hours total, under good conditions. Many patients can’t have the injection at all — an estimated 13.4% of the global population has some form of chronic kidney disease. Others simply can’t afford the 60% premium on the base scan cost.

Gadolinium contrast is roughly a 1.5 billion USD per year market, with six major producers. The injection is the default of neuro- radiology because nothing better existed. That’s where we enter.

What we’re building

Pixelence uses a generative model (U-Net with attention blocks, as described in our filed patent) to synthesize contrast-enhanced brain images from standard, non-contrast MRI sequences. The model takes T1, T2, and FLAIR inputs that every modern MRI already produces, and generates the equivalent contrast-enhanced view — plus a mask highlighting the region of interest for the radiologist.

The result, measured on our clinical trial: in under 30 seconds, a radiologist receives a contrast-equivalent view with zero patient side effects, no second scan, and roughly 60% cost reduction at the scan level. For the hospital, it’s 2x the effective ROI on an MRI machine they already own.

As CPO, my work has covered the product end-to-end: the DICOM processing pipeline, the annotation strategy with our clinical partners, the evaluation protocol across three hospital sites, the hospital-side web interface and API, and the privacy architecture that lets us work with real patient data under regulated constraints.

Where it stands

Peer-reviewed validation. Our clinical study of 156 brain tumor patients — 40 doctor reviews, peer-reviewed by 9 researchers, conducted with Sohag Medical Centre in Egypt — was published in the Journal of Clinical Oncology.

Active clinical trials at NUHS, IDC, and UPM in Malaysia. Registered with Malaysia’s National Medical Research Register (NMRR). Training data partnerships covering 350K patient records (Elborg Radiology Centre) and 17M EMR records (Islamabad Diagnostic Centre).

Funding. RM 200K raised to date (Cradle CIP Spark Grant and WatchTower VC). European Innovation Council Grant Phase 1 approved, with a €2.5M Phase 2 target. Currently raising a USD 800K round against a clear milestone plan: CE and FDA pathway, 95% accuracy target, and sales infrastructure.

Recognition. 2nd place at Eureka IIEC 2025 in Zagreb (60+ global applicants). Top 3 at Malaysia Startup World Cup 2024. Finalist at Startup World Cup West Malaysia 2025. Top 6 at Selangor Deep Tech Xccelerator. Top 3 at NEXEA Selangor Life Sciences Accelerator. Selected for the Cyberview Living Lab Accelerator Cohort 19.

Recent milestones. Patent filed in March 2026 for our Gadolinium-free contrast MRI synthesis. Exhibited and pitched at SusHi Tech Tokyo 2026. Selected to pitch at the 2026 SelectUSA Investment Summit in Maryland. Full news timeline →

The team

I came into Antler with the conviction that preventive healthcare was where I wanted to build. The brain MRI direction, the co-founder search, the decision to bring in a clinical oncologist as our third founder — those were the earliest moves. But Pixelence as a company is a three-person effort, and it has been from the start.

Raheel Zubairi leads as CEO. His relentless push — through the fundraising rounds, the grant applications, the accelerator cycles, the international pitches — is a large part of why Pixelence is where it is today. His background spans fintech, governance, and digital transformation. Amr Mohammad leads as CTO and CMO. He’s a clinical oncologist and data scientist, formerly Ex-Radiotherapy Chief in Egypt, currently with the NHS, and his published research in this specific domain was the reason the technology had a foundation to stand on from day one.

I lead product as CPO. We’re supported by clinical partners in Malaysia (Khairil Aswad Othman as Principal Investigator, Dayang Jok as Clinical Research Consultant) and a four-person advisory board including a site chief radiologist, two MDs, and a PhD in AI/ML and NLP.