Analytics based on artificial intelligence has advanced scientific research fields like natural language processing and imaging classification. Clinical research has also been influenced by artificial intelligence. Renal dysfunctions, presented as acute kidney injury (AKI) in critical care condition or chronic kidney diseases, are important conditions for patients, which require accurate risk stratification and initiation of rescue therapy. Thus, artificial intelligence can have profound impact on the management of these patients. Furthermore, patients with risk factors for AKI such as those with sepsis, acute respiratory distress syndrome and trauma are comprised of heterogeneous population. The “one-size-fit-all” paradigm may not fit for the management of such a heterogeneous patient population. Thus, tools from artificial intelligence can be employed to identify novel sub-phenotypes of these patients. These sub-classifications can not only provide prognostic value for risk stratification, but also predictive value for individualized treatment. Transcriptome can provide substantial amounts of information for a patient; artificial intelligence can help to identify useful information from such high dimensional data.
The issue primarily focuses on the use of artificial intelligence for the diagnosis and treatment of patients with or at risk for renal dysfunctions. A large amount of data is being generated from electronic healthcare records and transcriptome analysis. Novel methods from artificial intelligence can help to address the curse of dimensionality as have been frequently encountered when substantial number of variables are being processed with conventional methods. This focused issue also welcomes submissions of bioinformatic analysis with methods such as deep learning, density estimation and reinforcement learning. In such a way, these advanced machine learning methods can help to provide novel findings from large amount of data. Traditional methods in the context of epidemiology and medical statistics may fail to provide such novel findings due to their intrinsic limitations. Studies focusing on the topics below are welcome:
- Predictive analytics for risk stratification of patients with or at risk for renal dysfunctions
- Individualized treatment strategy for patients with rapidly changing conditions
- Sub phenotypes of heterogeneous population with or at risk for renal dysfunctions
- Bioinformatics analysis with transcriptome to develop individualized management
