While I write often about the limitations of artificial intelligence and what it can really offer in practice today, sometimes I come across developments in A.I. that not only blow my mind away but also depict an exciting vision of the future of healthcare. Look at this: In some cases, patients need injections with an iodine-based dye (radiocontrast) to make their organs more visible in scans. But some patients can’t take the dye. So researchers developed a GAN (generative adversarial network) that generates labeled, undyed CT images of lesions on kidneys, spleens, and livers. They added these images to real-world training data to improve performance of a segmentation model that marks lesions in diagnostic scans. As it's really hard to get enough quality medical data to train the algorithm on and labeled medical data is even harder to find, GANs are making it easier and less expensive to create large, annotated datasets for training A.I. diagnostic tools. #artificialintelligence #ai #aihealthcare #digitalhealth #futureofhealthcare #themedicalfuturist …see more