News

Estonian MedTech Start-Up Pioneers AI Breakthrough in Cancer Diagnosis

Published on: 4 November 2025
Member News
AI, Cancer
Better Medicine team II

Medtech company Better Medicine a member of our Estonian Tehnopol HealthTech Ecosystem, aims to become a world’s leading provider of intelligent cancer diagnostics. The company is currently focusing on abdominal tumors, which are among challenging to detect. Their product BMVision Kidney is the first AI-based solution for kidney cancer detection to receive CE certification under the EU Medical Device Regulation (MDR 2017/745).

Cancer cases are on the rise worldwide, and so are the costs of treatment. Kidney cancer, the 14th most common cancer globally, accounts for around 2.2% of all new cancer cases each year, with over 430,000 new diagnoses and nearly 156,000 deaths annually. Although often considered relatively rare, it ranks 16th in cancer-related mortality.

Beyond its human impact, kidney cancer represents a significant and growing economic burden. Between 2020 and 2050, the global cost of kidney cancer is projected to reach $536 billion, equivalent to about 0.012% of global GDP.

As both incidence and costs continue to climb, kidney cancer is emerging as a major health and economic challenge – underscoring the need for earlier detection, smarter diagnostic tools, and more efficient care pathways.

The power of smarter cancer diagnostics

With healthcare costs rising globally, the pressure on medical systems to deliver more efficient and accurate care has never been greater. In this context, modern technology plays a crucial role, not only optimizing resources but also serving as a trusted “second pair of eyes” for clinicians.

Founded in 2020, Better Medicine was born from a growing global demand for medical technologies helping to detect cancer in early phases. What started as an ambition to enhance cancer diagnostics evolved into years of research and scientific collaboration with universities and culminating in BMVision Kidney, the first AI-based kidney cancer detection technology to receive CE certification under the EU Medical Device Regulation (MDR 2017/745).

Recognizing that saving lives and reducing the cost of cancer care both start with early detection, Better Medicine set out to empower radiologists whose shortage is not just Estonian national concern, but a global one.

And the results speak for themselves: in clinical studies, radiologists using BMVision achieved a 99.2% sensitivity in detecting patients with kidney cancer, underscoring its potential to transform diagnostic accuracy.

Innovation is no longer optional in healthcare, it’s essential. The medical sector must continue to evolve, not only to save lives but also to enhance efficiency and ensure financial sustainability. This transformation requires agility: the ability to adopt and integrate new technologies quickly while maintaining the highest standards of patient privacy, data security, and ethical responsibility. 

By fostering a culture of adaptability and trust in innovation, healthcare can move faster toward a future where technology and human expertise work seamlessly together to deliver better outcomes for all.

Better Medicine Explores New Frontiers in Radiology Innovation

Better Medicine is discovering new ways to bring innovation into radiology and improve cancer diagnostics. The company’s latest success came at the MICCAI Society CLIP Workshop 2025, where its research received the Best Paper Award.

The paper, written by David Avedis Injarabian, Joonas Ariva, Dmytro Fishman and Hendrik Šuvalov, began as a master’s thesis. It explored how Contrastive Language-Image Pretraining (CLIP) models can be used in medical imaging. The main question was whether AI systems could learn directly from radiology reports and still identify and segment disease patterns on CT scans.

This idea could change how medical AI is built. Manual segmentation maps are expensive and time-consuming to produce. Radiologists, however, already create detailed reports every day. If AI could learn from these reports, training models would become much faster and more affordable.

The researchers developed the CT-CLIP model as a proof of concept. It showed that textual and visual data can be combined in meaningful ways. The study also revealed that current AI tools are not yet fully ready for this approach. Further work is needed to improve the technology and bring it into clinical use.

This research shows we’re on the right track – using textual radiology reports to guide medical AI opens up promising possibilities. We’re excited to push this research further and develop systems that can detect pathology regions from reports, making medical AI training more scalable and efficient.

Healthcare is changing fast, and success depends on how quickly institutions embrace innovation. Better Medicine aims to lead in intelligent cancer diagnostics, expanding across Europe and into the US. In partnership with top hospitals and universities, the company advances early detection and builds a more efficient, patient-focused healthcare future.