Uncovering Small Secrets in Big Data Sets
How Math Can Identify Biology in Rare Conditions (Pediatric Pulmonary Hypertension)
This article requires a subscription to view the full text. If you have a subscription you may use the login form below to view the article. Access to this article can also be purchased.
Complex patterns recognition is an evolutionary capability that differentiates humans from any other animal. Whereas environment recognition and emotional state changes are seen in many different mammals, the cognitive repertoire of humans is more comprehensive and multifaceted.1 Historically, progressive development in medicine has been somehow linked to this particular ability. Clinical syndromes, diagnostic tools, and treatments have been developed based on classification systems. Classification systems have also been incorporated in the clinical practice by our ability to group similar patients according to symptoms, pathological findings, and pathophysiological aspects of any determined condition.2 Classification systems are dynamic tools that incorporate advancing technologies and allow better communication among physicians and researchers in prediction of treatment response and prognosis.3
Article, see p 341
Because biotechnology has improved, our ability to phenotype disease with genomics, proteomics, and imaging modalities is now sophisticated. Usual statistical analysis lacks the discriminative ability to handle so many factors. Advanced statistical modeling and artificial intelligence are the next step in advancing medical science to handle all the data and lead to a more scientific phenotyping of disease. One such method is neuronal networks. Like the brain, many factors (nodes) are interlinked in multiple circuits to form a network. This allows for multiple interactions to occur with multiple linkage combinations. Networks provide information on biological processes and clinical phenotypic differences between patients with different demographics, disease progression rates, and mortalities.4
Pulmonary hypertension (PH) is a heterogeneous disease with its current classification schema based on commonalities in cause, pathology, and expert opinion. PH is currently classified into 5 groups: group 1 pulmonary arterial hypertension, group 2 PH associated with left heart disease, group 3 …