Variant spectrum of F8 and F9 in hemophilia patients from southern China and 26 novel variants.

Hemophilia, an X-linked recessive disorder, is characterized by spontaneous or trauma-induced prolonged bleeding. It is classified as hemophilia A when caused by variants in the F8 gene, and hemophilia B when caused by F9 variants. Few studies have described hemophilia variants in the Chinese population. This study aimed to investigate the clinical and genetic profiles of 193 hemophilia patients from southern China. Utilizing Sanger sequencing, multiplex ligation-dependent probe amplification, gap detection, long-range PCR, and multiplex PCR, we identified both F8 and F9 gene variants. Pregnant women with a history of hemophilia A offspring underwent amniocentesis or villus sampling for the variant detection. Variants in F8 and F9 were pinpointed in 183 patients, with 26 being novel discoveries. Notably, genetic testing was absent in the initial evaluation of 133 out of 161 patients, leading to a protracted average definitive diagnosis timeline of 2 years. Remarkably, two hemophilia A cases with anticipated severe phenotypes due to protein-truncating variants presented with only moderate or mild clinical manifestations. Among the 40 fetuses tested, 34 were males, with 17 exhibiting hemizygous variants in the F8 gene. Our results contribute to the broader understanding of F8 and F9 variant spectrum and highlight the underuse of genetic analyses in southern China.

Evaluation and accurate diagnoses of pediatric diseases using artificial intelligence.

Artificial intelligence (AI)-based methods have emerged as powerful tools to transform medical care. Although machine learning classifiers (MLCs) have already demonstrated strong performance in image-based diagnoses, analysis of diverse and massive electronic health record (EHR) data remains challenging. Here, we show that MLCs can query EHRs in a manner similar to the hypothetico-deductive reasoning used by physicians and unearth associations that previous statistical methods have not found. Our model applies an automated natural language processing system using deep learning techniques to extract clinically relevant information from EHRs. In total, 101.6 million data points from 1,362,559 pediatric patient visits presenting to a major referral center were analyzed to train and validate the framework. Our model demonstrates high diagnostic accuracy across multiple organ systems and is comparable to experienced pediatricians in diagnosing common childhood diseases. Our study provides a proof of concept for implementing an AI-based system as a means to aid physicians in tackling large amounts of data, augmenting diagnostic evaluations, and to provide clinical decision support in cases of diagnostic uncertainty or complexity. Although this impact may be most evident in areas where healthcare providers are in relative shortage, the benefits of such an AI system are likely to be universal.