CancerCenter.AI transforms cancer diagnostics, cutting diagnosis times with the help of Azure-powered platform

When Piotr Krajewski’s six-year-old nephew Jakob was diagnosed with brain cancer, the family’s ordeal exposed weaknesses in cancer diagnostics. Jakob endured a month-long wait for an initial diagnosis that led to immediate chemotherapy, only to learn after another month and a second opinion that he had been misdiagnosed. “I experienced firsthand how misdiagnosis and prolonged waiting times cause immense stress for patients and their families,” says Krajewski. “That’s why I founded CancerCenter.AI—with the mission to save lives by expediting and improving cancer diagnostics.”

Identifying systemic challenges in diagnostic medicine

Krajewski knew that the issue lay with pathologists and their outdated tools. "Many pathologists are still using microscopes that haven’t changed in over a century instead of relying on modern digital imaging technologies and digital patient data," he notes. This causes delays in analysis. The issue is exacerbated by a lack of specialists. “In Poland, we only have about 400 active pathologists,” notes Dr. Artur Bartczak, Pathologist at Mazovia Hospital. “Each one focuses on a type of cancer. For instance, only five pathologists focus on pediatric brain tumors.”

Additional challenges include the time-consuming process of sending physical tissue samples for a second opinion, which can lead to further delays. “Waiting for physical samples to be transported between hospitals can take weeks,” Krajewski explains. 

Selecting the right technology partner

With the number of cancer cases increasing annually, CancerCenter.AI’s primary challenge was to develop a robust, scalable platform that could process and analyze vast amounts of pathological data quickly and accurately. “Our goal was not only to speed up the diagnosis process but also increase its accuracy using AI—potentially saving lives through earlier detection and more precise treatment planning,” says Krajewski.

Krajewski needed a technology partner and solution that could meet specific requirements. “We chose Microsoft Azure for several key reasons. First was its versatile suite of data processing and AI tools, with advanced machine learning capabilities,” says Krajewski. “Another factor was compliance with our data residency requirements. With Azure Blob Storage, we can ensure that sensitive medical data remains within the customer’s country. The flexibility of Azure means we can also support all three options a customer might need: hybrid, on-premises, or cloud implementations."

The strong support Microsoft provides to startups also helped. "We joined Microsoft for Startups program and received technological support from Microsoft, enabling us to accelerate our development efforts," Krajewski explains.

Finally, the Azure Marketplace offered the benefit of global reach. "Microsoft has many customers in the healthcare space looking for products they can use together with their Azure services. So, we applied to become a Microsoft Partner and listed our platform on the Azure Marketplace," says Krajewski. "This visibility enables us to reach customers that otherwise might not be able to know we exist."

Empowering physicians with integrated cancer diagnostics

With this foundation, Krajewski and his team launched a comprehensive platform that supports physicians in both pathology and radiology. 

To overcome the lack of digital data, CancerCenter.AI developed the PathoCam software. PathoCam enables pathologists to create digital slides from their microscope cameras by capturing and stitching together multiple snapshots, then uploading them to the online platform for analysis. This offers an affordable alternative to expensive scanners or for urgent (intraoperative) cases. The platform also integrates with any type of scanner. “The platform offers magnification to analyze cells and nuclei,” explains Krajewski. Physicians can zoom in, annotate, analyze with AI, and share cases or images through their browsers. 

CancerCenter.AI also developed a radiology module which enables physicians to open and analyze radiology images on any device, from any location. This extends physicians’ ability to collaborate with experts and specialists anywhere in the world. Unlike legacy systems, CancerCenter.AI integrates pathology and radiology into a single platform, streamlining multidisciplinary collaboration.

To further facilitate cancer diagnosis, CancerCenter.AI trained AI models to review digital images. For one, the software assists physicians in identifying regions of interest. “An AI model may scan magnetic resonance imaging (MRI) images and quickly locate the prostate. This helps doctors retrieve the right tissue during a biopsy,” explains Krajewski. The AI models can also measure critical markers—such as the presence and intensity of receptors within tissue samples to classify based on Gleason grading system—to aid in cancer diagnosis.

Moreover, CancerCenter.AI added a Laboratory Information System module to enhance operational efficiency. "Our system transforms laboratories from paper-heavy environments into efficient diagnostic centers. We offer an intuitive interface and automation capabilities that eliminate frustrating procedures and chaotic specimen management, developed in close collaboration with end-users—the system adapts to your laboratory, not the other way around,” Krajewski explains. “Laboratory directors gain complete control over the diagnostic process, knowing exactly where each sample is located, who's responsible for the examination, and when results will be ready. This module increases laboratory throughput, directly translating to cost optimization and faster delivery of potentially life-saving results to patients, all while ensuring the system evolves based on real user feedback and needs." 

Quicker diagnosis processes, while reducing error 

"This platform has transformed our workflow," assures Dr. Bartczak, since pathologists can now preview and work on digital files on their phones and from anywhere. “This functionality is crucial because a lot of pathologists often work in multiple hospitals,” Krajewski explains. “When a clinic needs them to take a look at a new or urgent case and the pathologist is not in the building, our platform sends out an automatic notification. The pathologist can access the imaging through their phone for a quick diagnosis.”

This ability to check digital images on any device also improves physicians' work-life balance. "One pathologist told me that he didn’t have time to take any holidays because he is the only pathologist in his laboratory. Once the hospital started using our platform, he was finally able to go on holiday because he could examine urgent cases on the go," shares Krajewski. 

Today, thanks to AI-powered tools provided by CancerCenter.AI’s platform, pathologists can process cases and images faster, delivering final results to worried patients and families more quickly. "Traditionally, it may take seven to 15 minutes to manually analyze prostate cancer cases. With the help of our Azure-powered platform and AI models, this can be reduced to about a few minutes," reports Krajewski. 

Getting a second opinion also takes less time. "Sending real tissue to another clinic can take days or weeks," says Krajewski. "With our platform, doctors can exchange digital files, resulting in a notable acceleration of the diagnostic process and enhanced consultation and second opinion processes.” 

While specific numbers may vary depending on the healthcare setting, CancerCenter.AI observed great results. “With our Azure-based solution, we’ve helped accelerate cancer diagnosis processes and reduce potential diagnostic errors. Patients can receive critical results faster, enabling earlier treatment,” shares Krajewski.

Looking back, Krajewski is happy with his decision to choose Microsoft solutions as the foundation for CancerCenter.AI’s platform. “When we started using Azure Machine Learning to train new AI models, we used Azure Virtual Machines to provide new team members with the computational resources they needed without investing in physical hardware,” shares Jaroslaw Kwiecien, CTO at CancerCenter.AI. “As a result, we have onboarded more data scientists and trained AI models faster.” Azure Kubernetes Service also helped effectively manage and scale AI models.

Expanding the mission globally

Jakob’s medical journey has inspired Krajewski to pave the way for the next generation of patients. "Today, our platform is used by 17 hospitals, not only for gathering second opinions, but also as the primary digital pathology tool. Most are in Poland, one in Switzerland, and one in Georgia. We are also working on expanding into Africa, Americas, Asia, and Central and Eastern Europe, where there aren’t enough professionals. Supporting them is a key focus for us," notes Krajewski.

Krajewski firmly believes the key to minimizing wait times for patients lies in solving the challenges physicians face. Given the global shortage of physicians, the power of AI can truly make impactful changes. That’s why Krajewski aims to make pathology more accessible by offering the platform pro bono for educational purposes. Already, pathologists use it to showcase new cases during conferences. He says, “By promoting digital pathology, telepathology solutions, and AI within the medical community, we believe we can truly make a difference in the lives of patients and their families.”

Discover more about CancerCenter.AI on its website or via social media: LinkedIn, Facebook, X/Twitter, and YouTube.