Artifical Intelligence - preliminary models
Will there come a day when artificial intelligence (AI) will change the way we study and understand biological structures like bones and tissues ? Can we, on the dawn of AI revolution, predict its potential in, for example, routine medical imaging data segmentation used for 3D printing ? AI is making waves in medical imaging, believing to faster and more precise identify and segment these structures. But at what stage are we right now ?
When participating on ACCEDE project, we were approached by slovak company Medannot (https://medannot.com/), the first end-to-end, no-code automation platform designed to revolutionize radiology workflows. Supporting a wide range of imaging modalities, including CT, MRI, RTG, and ECHO, Medannot enables healthcare professionals to work with any DICOM image of any body part—all within a single, browser-based platform. With Medannot, clinicians can seamlessly pull images from their PACS, annotate them, develop AI models, train and clinically validate these models, and deploy them as workflows directly integrated into their PACS. All of this is achieved without writing a single line of code. The workflows created can serve diagnostic purposes, such as kidney tumor screening, 3D modeling for preoperative planning, or even therapeutic applications like monitoring tumor size post-treatment.
What was important for us, Medannot enables the creation of 3D models from AI-annotated and segmented images. These patient-specific 3D models are invaluable for education, preoperative planning, and research. Surgeons can visualize anatomical structures with unparalleled clarity, making complex procedures safer and more efficient. For the sake of our project, we were provided 16 biological structures segmented by Medannot´s AI system. When thoroughly checked, they still differ from the standard models, they lack a lot of details but the outcome is far different - we can see how AI learns, we can see in practically real-time how the quality is improving with more and more structures segmented.
Radiology is vast and repetitive, making automation essential—but no single solution can address every need. Medannot stands apart as an infrastructure company empowering clinicians worldwide with the tools to create their own custom AI-powered solutions. By enabling personalized, automated workflows and advanced modeling capabilities, Medannot is reshaping the future of radiology and advancing global healthcare innovation.
Thus, we presented, for you to enjoy, several nice models created by Medannot´s AI. Feel free to use and let us know your opinion !