With respect to the concept of generative artificial intelligence, we conducted some experiment in order to try to accelerate both translational research in MDS as well as clinical trials. In particular, there is an emerging technology that is the generation of the so-called synthetic data that can really help us to accelerate the generation of evidence for patients with myelodysplastic syndromes...
With respect to the concept of generative artificial intelligence, we conducted some experiment in order to try to accelerate both translational research in MDS as well as clinical trials. In particular, there is an emerging technology that is the generation of the so-called synthetic data that can really help us to accelerate the generation of evidence for patients with myelodysplastic syndromes. For instance, in more practical terms, we know that in in MDS, the first evidence for the definition of recurrent genetic mutations in a patient affected with this disorder was published in 2011 and only ten years later, a large patient population mean more than 2000 patients with comprehensive clinical and genomic information were available, who provide a new score for patient prognostication. So, this means that the development of innovative data driven digital health products or solution is slowed due to the limited access to the clinical data and additional challenges are expected in data collection in rare diseases such as myelodysplastic syndromes. In this context, there is an emerging technology the generation of synthetic data that are artificial, the artificial data that are generated by a specific algorithm that is trained to learn all the essential statistical and clinical characteristics of a real data set. These new data are not a copy or representation of the real data and according to this definition, synthetic data are not regulated by particular limitation and can be easily accessed and shared. So, this technology can accelerate the translational research and the clinical trials by allowing data sharing in a way that is totally compliant with privacy preservability to solve some issues in data collection of real information such as classes, imbalance, and missing information. We can provide data augmentation so we can produce a dataset of 5000 patients starting from one or a few hundred patients and is able to generate new ones. So, this is a practical example of the use of generative artificial intelligence to try to accelerate the clinical research in patients with rare hematological neoplasms.