AI Connecting Cancer Scans to Care
By Nina Karmelitow— working in Strategic Communication and Investor Relations within biotech and medtech.
August 20, 2025
New AI tools are transforming cancer care. Beyond detecting disease, they’re now guiding what happens next by turning scans into tailored care plans. Successful implementation involves software installation, secure connections to patient data, and staff training to ensure safety. In well-equipped rural clinics, they enable rapid scanning, risk assessment, and treatment planning before patients leave—resulting in faster care, fewer unnecessary steps, and lower costs. This helps reduce delays that worsen disease and burden healthcare systems, ultimately benefiting patients and the wider community.
Swedish gynecological diagnostics company Intelligyn exemplifies this shift. Their patent pending transformer-based AI, published in Nature Medicine, does more than diagnose ovarian tumors from ultrasound; it organizes tumor types into a detailed hierarchy so it can distinguish subtle differences between many tumor subtypes. The system can automatically remove irrelevant artifacts from images to focus on important diagnostic features. Validated on over 17,000 images from more than 3,600 patients across 20 international centers, this design allows for seamless integration into existing hospital systems, turning a diagnostic scan into a treatment pathway in real time.
Global Leaders and Their Timelines
Other companies are racing toward similar milestones. In the U.S., Paige, backed by $241 million, injects AI-driven pathology into treatment workflows. Lunit, a South Korean company specializing in methods and systems for analyzing medical images, is now valued at approximately $1.1 billion and earned approximately USD 15 million in revenue during Q1 2025 representing a 273.6% increase year-over-year compared to Q1 2024. Meanwhile, India’s Qure.ai, recognized among TIME’s 100 Most Influential Companies of 2025, has developed AI-powered diagnostic tools for medical imaging that are active in over 100 countries. The company has pulled in $125 million in combined funding and revenue, with its systems deployed live in under two years. Global adoption is accelerating, but as these systems move from research to real-world deployment, the question becomes not just how fast we can implement them, but how safely.
Safety First, Always
For cancer treatment planning, integrating robust diagnostic artificial intelligence is becoming the norm. Hospitals expect partners to provide tools supported by regular audits, validation, and human oversight. This reflects the lifecycle mindset promoted by the US FDA, which calls for new guidance on Artificial Intelligence-Enabled Device Software. To meet this vision, machine-learning–enabled medical devices must be built on transparency and fairness, demonstrating that these tools serve all patient groups equally well, with ongoing monitoring for errors and prompt corrections when needed. Success depends on establishing a strong foundation and involving clinicians in design and deployment. Those who act now can earn trust, raise the standard, and lead as healthcare moves toward fully integrated, intelligence-driven workflows.