%0 Journal Article %1 Eckardt2024-jt %A Eckardt, Jan-Niklas %A Hahn, Waldemar %A Röllig, Christoph %A Stasik, Sebastian %A Platzbecker, Uwe %A Müller-Tidow, Carsten %A Serve, Hubert %A Baldus, Claudia D %A Schliemann, Christoph %A Schäfer-Eckart, Kerstin %A Hanoun, Maher %A Kaufmann, Martin %A Burchert, Andreas %A Thiede, Christian %A Schetelig, Johannes %A Sedlmayr, Martin %A Bornhäuser, Martin %A Wolfien, Markus %A Middeke, Jan Moritz %D 2024 %J NPJ Digit. Med. %K ai artificial clinical generative intelligence leukemia myeloid synthetic trials yaff %N 1 %P 76 %T Mimicking clinical trials with synthetic acute myeloid leukemia patients using generative artificial intelligence %V 7 %X Clinical research relies on high-quality patient data, however, obtaining big data sets is costly and access to existing data is often hindered by privacy and regulatory concerns. Synthetic data generation holds the promise of effectively bypassing these boundaries allowing for simplified data accessibility and the prospect of synthetic control cohorts. We employed two different methodologies of generative artificial intelligence - CTAB-GAN+ and normalizing flows (NFlow) - to synthesize patient data derived from 1606 patients with acute myeloid leukemia, a heterogeneous hematological malignancy, that were treated within four multicenter clinical trials. Both generative models accurately captured distributions of demographic, laboratory, molecular and cytogenetic variables, as well as patient outcomes yielding high performance scores regarding fidelity and usability of both synthetic cohorts (n = 1606 each). Survival analysis demonstrated close resemblance of survival curves between original and synthetic cohorts. Inter-variable relationships were preserved in univariable outcome analysis enabling explorative analysis in our synthetic data. Additionally, training sample privacy is safeguarded mitigating possible patient re-identification, which we quantified using Hamming distances. We provide not only a proof-of-concept for synthetic data generation in multimodal clinical data for rare diseases, but also full public access to synthetic data sets to foster further research.

@article{Eckardt2024-jt, abstract = {Clinical research relies on high-quality patient data, however, obtaining big data sets is costly and access to existing data is often hindered by privacy and regulatory concerns. Synthetic data generation holds the promise of effectively bypassing these boundaries allowing for simplified data accessibility and the prospect of synthetic control cohorts. We employed two different methodologies of generative artificial intelligence - CTAB-GAN+ and normalizing flows (NFlow) - to synthesize patient data derived from 1606 patients with acute myeloid leukemia, a heterogeneous hematological malignancy, that were treated within four multicenter clinical trials. Both generative models accurately captured distributions of demographic, laboratory, molecular and cytogenetic variables, as well as patient outcomes yielding high performance scores regarding fidelity and usability of both synthetic cohorts (n = 1606 each). Survival analysis demonstrated close resemblance of survival curves between original and synthetic cohorts. Inter-variable relationships were preserved in univariable outcome analysis enabling explorative analysis in our synthetic data. Additionally, training sample privacy is safeguarded mitigating possible patient re-identification, which we quantified using Hamming distances. We provide not only a proof-of-concept for synthetic data generation in multimodal clinical data for rare diseases, but also full public access to synthetic data sets to foster further research.}, added-at = {2025-01-07T16:33:43.000+0100}, author = {Eckardt, Jan-Niklas and Hahn, Waldemar and R{\"o}llig, Christoph and Stasik, Sebastian and Platzbecker, Uwe and M{\"u}ller-Tidow, Carsten and Serve, Hubert and Baldus, Claudia D and Schliemann, Christoph and Sch{\"a}fer-Eckart, Kerstin and Hanoun, Maher and Kaufmann, Martin and Burchert, Andreas and Thiede, Christian and Schetelig, Johannes and Sedlmayr, Martin and Bornh{\"a}user, Martin and Wolfien, Markus and Middeke, Jan Moritz}, biburl = {https://puma.scadsai.uni-leipzig.de/bibtex/2abf30651406b6b4803daa9239832f9a2/scadsfct}, interhash = {612c8ed14c52ea836d2a28d9254624cb}, intrahash = {abf30651406b6b4803daa9239832f9a2}, journal = {NPJ Digit. Med.}, keywords = {ai artificial clinical generative intelligence leukemia myeloid synthetic trials yaff}, language = {en}, month = mar, number = 1, pages = 76, timestamp = {2025-07-29T10:49:33.000+0200}, title = {Mimicking clinical trials with synthetic acute myeloid leukemia patients using generative artificial intelligence}, volume = 7, year = 2024 }