From the slide rule to the calculator is only a step—and healthcare is experiencing a similar shift today with artificial intelligence. AI in medical imaging is not a fad, but a longstanding field standing on the shoulders of university hospitals. Its promise lies in more accurate diagnostics, a lower risk of missed findings, and faster care.
Three pillars: radiology, radiotherapy, and pathology
In radiology, AI helps uncover malignant lesions, microfractures, and other subtle findings, reducing the risk of oversight and saving time. A real-world experience of waiting a week for an MRI result shows why automation is so necessary. In radiotherapy, AI is used mainly to segment organs at risk for treatment planning, with an eye toward broader involvement in the planning itself. The goal is safer, more precise treatment with less burden on both the patient and the team.
In pathology, AI helps identify histopathological findings more quickly and precisely, for example in prostate carcinoma, one of the most common in men over fifty. On its own, AI today often performs at the level of a junior specialist, but in tandem with one it achieves senior-level accuracy. When a senior expert joins forces with AI, the difference compared to a junior is immense. An important task is also assessing whether the tumor extends beyond the boundaries of the prostate, which fundamentally influences the surgical strategy—this is exactly where AI can contribute to more precise decision-making.
From data to practice: workflow and collaborations
A disciplined process underpins successful prediction: collecting high-quality data, expert annotation, preprocessing, training, and validation of models. Only then comes the “wrapping” into a user interface and deployment, where inference runs on the server in real time. Such a workflow ensures that the tool is not only accurate in a study but also usable in a hospital. Emphasis on transparency and validation is key to physicians’ trust.
Practical projects show what is possible: for example, a model for aneurysm detection was developed in collaboration with Univerzitnou nemocnicou v Martine, others focus on prostate cancer, head and neck tumors, or lymphomas. Behind these initiatives are people who brought know-how from abroad back home to connect laboratory research with everyday medicine. The ambition is to build a bridge between theory and practice—as seamless as the transition from the slide rule to the calculator.
