Skip to main content


AI-Assisted Cancer Detection Speeds Diagnosis

AI assisted

The most transformative innovations often emerge during times of crisis. That was the case for Javier García López and his cofounders at Sycai Technologies, a Barcelona-based digital health company.

The company launched in February 2020—one month before the pandemic—intending to create a marketplace where users would download AI-trained models created for their own relevant use case. But the pandemic created a more urgent need for health systems and providers to leverage technology to enable their work. As a result, the global health crisis unwittingly gave Sycai a powerful medical need for its solution and opened a pathway for the company to explore partnerships with hospitals to test its application.

During this time, Sycai also discovered an even more urgent use for its technology related to pancreatic diseases. Today, its original application has evolved into Sycai Medical, an AI assistant that uses machine learning and neural networks to empower radiologists to more accurately detect and diagnose pancreatic cancer.

Confronting a Silent Pandemic

Pancreatic cancer is usually diagnosed at a late stage and has one of the lowest 5-year survival rates among cancers. But Sycai Medical is harnessing AI to address this challenge, with a solution that can detect precancerous lesions in the upper abdomen much earlier—and improve cancer care if the disease is diagnosed.

“It was something that everyone told us is really dangerous. It’s like a silent pandemic. Up to one fourth of the population has these kinds of lesions, but they’re never detected on time because they have no prior symptoms. So, we thought we could have a chance if we were to focus there,” says García López, Sycai’s chief technology officer and cofounder.

García López, CTO, founded Sycai Technologies along with Sara Toledano, the company’s CEO. They then met their third cofounder Júlia Rodríguez Comas, who now serves as Sycai’s chief scientific officer. Comas, a scientist and researcher, has a Ph.D. in biomedicine and specializes in the pancreas. Her clinical knowledge propelled the team to focus on the pancreas and address this longstanding challenge in the medical field.

Within radiology, AI typically is applied to brain, lung, and breast conditions, García López says. The pancreas largely has been uncharted territory, but Sycai Medical may change this. With the help of application programming interfaces (APIs), the solution easily integrates into a hospital’s existing medical imaging system.

Sycai Medical reprocesses and analyzes a patient’s scan and then normalizes the image, so all the organs are equally visible on the scan. Next, neural networks (AI models) trained on anonymized data from thousands of patients with lesions in the upper abdomen pinpoint the exact location of the pancreas within the abdomen. Once AI identifies the pancreas’ location, it determines whether a lesion is present, and if so, whether its composition and characteristics indicate it is cancerous, precancerous, or benign.

“It extracts multiple parameters that if you match them with the clinical guidelines, it finally gives you what is the malignancy potential of these lesions,” García López says.

#AI works quietly in the background to surface this valuable information—without interrupting radiologists’ typical #workflow. @SycaiT via @insightdottech

AI works quietly in the background to surface this valuable information—without interrupting radiologists’ typical workflow. Sycai Medical complements their work without making a final clinical judgment for them. García López says the tool acts as a diagnostic assistant, warning doctors that it has found something on the scan that could be dangerous to the patient. Doctors can choose to open the alert and investigate further at their own discretion. The GDPR-compliant solution also doesn’t capture any metadata that could identify patients and is designed to ensure there are no memory or data leaks once it integrates with a hospital’s IT system, even in case of a server attack.

Bringing AI-Assisted Cancer Detection to More Hospitals

Sycai Medical accelerates AI-assisted cancer detection using a range of technologies, including the Intel® OpenVINO toolkit, open-source software that deploys and optimizes the performance of AI models.

With OpenVINO, the software’s AI models have been able to diagnose a lesion’s potential malignancy 70% faster with less than a 3% impact on diagnostic accuracy. “We were still over 90% accurate with 70% less inference time,” García López says.

Sycai Medical is a powerful tool for accurate early detection of pancreatic cancer, but it also could prevent unnecessary biopsies if a lesion is benign, and optimize care management if a patient is diagnosed with a disease.

The company conducted clinical pilots of Sycai Medical with hospitals in Spain and Germany. It is now going through the regulatory process, where regulators will audit its previous clinical trials. The company plans to launch in Europe this year, with a focus on detecting and diagnosing pancreatic cystic lesions. The solution also has future implications for the early detection of other pathologies, such as liver and kidney disease, with hospitals testing this use case as well, García López says.

Healthcare providers in the U.S. may soon have access to the tool. Sycai Medical is currently undergoing a 6-month pilot test at the University of Alabama in advance of potential FDA approval.

The Sycai solution showcases the transformative power of AI and its role in supporting better healthcare outcomes. By harnessing AI to improve cancer detection, Sycai Medical is delivering insights that could innovate cancer treatment and care, empowering healthcare providers to diagnose diseases faster and more accurately—and potentially save many more lives.

This article was edited by Georganne Benesch, Editorial Director for

About the Author

Satta Sarmah Hightower is a journalist-turned-content marketer who produces content for agencies and brands in the healthcare, technology and financial services industries. Satta previously worked for the Orlando Sentinel and Patch Media, a division of AOL. At Patch, she was a reporter and editor before becoming the senior manager of editorial operations, where she oversaw national content sponsorships for Fortune 500 clients and well-known brands. Satta holds a bachelor's degree in journalism from Boston University and a master's degree in journalism from Northwestern University's Medill School.

Profile Photo of Satta Hightower