Imagine a world where cancer diagnoses are faster, more accurate, and accessible to everyone—regardless of where they live. That’s the future we’re building with HistoVault v1, Pakistan’s largest histopathology dataset, designed to supercharge AI-driven cancer research and diagnostics. And guess what? It’s just the beginning.
What is HistoVault v1?
Think of HistoVault v1 as a massive visual library of cancer images, built to train AI models in recognizing and diagnosing cancer with precision. With over 17,000 high-quality images, this dataset covers four major cancers: oral, breast, colorectal, and gastric, which were captured using two imaging techniques: Low-Cost Low-Resolution (LCLR) and Whole Slide Imaging (WSI).
In simple terms, it’s data that teaches AI how to spot cancer, making histopathology smarter, faster, and more reliable.
Why does this matter?
Cancer is a race against time. The earlier it’s caught, the better the chances of survival. But here’s the problem: traditional histopathology is time-consuming, heavily dependent on specialists, giving rise to the problem that there just aren’t enough trained professionals to keep up with rising cancer cases.
This is where AI can step in and help. HistoVault v1 is built to:
- Supercharge AI research in histopathology, allowing researchers to train and improve machine learning models for detecting cancer.
- Support histopathologists by automating tedious tasks, freeing up their time for complex cases.
- Make cancer diagnosis faster and more accessible, especially in regions with limited expertise.
- Boost collaboration by being open-source, giving researchers and developers worldwide the opportunity to build better, smarter diagnostic tools.
The team behind HistoVault v1
HistoVault v1 is the result of countless hours of dedication from an incredible team of researchers, histopathologists and data scientists.
The project was a powerhouse collaboration between the Precision Medicine Lab (PML) and the Department of Histopathology at Rehman Medical Institute (RMI). The effort was spearheaded by Dr. Faisal Khan (PI, PML) and Brig.(R) Prof. Dr. Iqbal Muhammad (Consultant & HOD, Histopathology, RMI), alongside a rock-solid team that kept everything moving smoothly and took on the Herculean task of capturing 17,000+ histopathology images—a job that demanded incredible focus and patience.
- PML Team: Dr. Madina Shirdel (Project Lead), Arsalan Riaz, Dr. Qazi Kamran Amin, Shaiq Paracha, Mudassir Shah, Muhammad Ahmad.
- Histopathologists: Dr. Maria Tasneem Khattak and her dedicated team of TMOs.
Launch day: A milestone for AI in medicine
HistoVault v1 was officially launched on 13 February 2025, which saw some of Pakistan’s leading minds in science, technology, and medicine come together. From policymakers to AI experts, everyone recognized the sheer potential of this dataset to redefine cancer diagnostics in Pakistan and beyond.
Among the distinguished guests were Mr. Sajid Hussain Shah (DG, Science & Technology, KP), Engr. Sohaib Tanveer (VP, CECOS University), Shafique Ur Rehman (CEO, RMI), and experts from Ghulam Ishaq Khan Institute of Technology, Institute of Management Sciences, including consultants from Khyber Teaching Hospital and Peshawar Medical College.
Words that stuck with us
“Just imagine a future where solutions developed right here in Hayatabad are saving lives—not just in Pakistan, but in London, Tokyo and New York.” – Shafique Ur Rehman, CEO, RMI
“AI in Medicine is here, and Pakistan needs to urgently scale up its capabilities. We couldn’t be more proud to be at the forefront of this.” – Dr. Faisal Khan, PI, PML
What’s next?
The beauty of Histovault is that it’s open-source! Soon, researchers and developers will be able to access this dataset via the Open Data Portal Pakistan, sparking new innovations in AI-driven histopathology.
From AI-assisted cancer screenings to automated histopathology reports, the possibilities are limitless. And this is just version 1, we’re only getting started!
