ARTIFICIAL INTELLIGENCE GENERATES SYNTHETIC DATA TO DISCRIMINATE PSORIATIC ARTHRITIS WITHOUT PSORIASIS FROM SERONEGATIVE RHEUMATOID ARTHRITIS
Keywords
artificial intelligence; synthetic data; data augmentation; machine learning
We studied 200 people with psoriatic arthritis without psoriasis or rheumatoid arthritis without autoantibodies-two conditions apparently similar with unclear discriminants. Using artificial intelligence, we created and validated “synthetic” patients resembling but not matching individuals from the real world. These increased the total patient number, making computer models better at telling the two diseases apart by weighing clinical parameters.


