A Visual Phenotype-Based Differential Diagnosis Process for Rare Diseases

Purpose Phenotype-based rapid diagnosis can make up for the time-consuming genetic sequencing diagnosis of rare diseases. However, the collected phenotypes of patients can sometimes be inaccurate or incomplete, which limits the accuracy of diagnostic results. To solve this problem, we try to design a phenotype-based differential diagnosis process for rare diseases to achieve rapid and accurate diagnosis of rare diseases. Methods The core of the differential diagnosis of rare diseases is to optimize the phenotype information of a specific patient and the visualized comparative analysis of diseases. To recommend additional phenotypes, replace the fuzzy phenotypes and filter the unexplained phenotypes for patients, we constructed a phenotype hierarchical network and a disease-phenotype differential network and calculated the phenotype co-occurrence relationship. In addition, we designed a visual comparative analysis method to explore the correlation and difference of disease phenotypes. Results The evaluation based on the published 10 rare disease cases demonstrated that after the optimization of patient phenotype information through our differential diagnosis, the target disease often got a better ranking and recommendation score than before. We have deployed this scheme on the RDmap project (http://rdmap.nbscn.org). Conclusion Compared to genetic and molecular analysis, phenotype-based diagnosis is faster, cheaper, and easier. The differential diagnosis process we designed can optimize the phenotype information of patients and better locate the target disease. It can also help to make screening decisions before genetic testing. Graphical Abstract Inaccurate, incomplete and mistaken phenotypes collected in the diagnosis of rare diseases will lead to misdiagnosis. This defect of artifically collected phenotypes is inevitable. [GRAPHICS] Imn response to this problem, we designes a rapid differential diagnosis process for are rare diseases based on phenotypic network, disease network and phenotypic co-occurrence relationship. Improve the recommended ranking of target diseases by optimizing collected phenotypes and visualized disease comparison analysis. [GRAPHICS]

Automated summary

This study proposes a phenotype-based differential diagnosis process for rare diseases, aiming to achieve rapid and accurate diagnosis by optimizing patient phenotype information. The process involves constructing a phenotype hierarchical network and a disease-phenotype differential network, calculating phenotype co-occurrence relationships, and designing a visual comparative analysis method to explore correlations and differences between disease phenotypes.