Population Health and Chronic Disease Management

Kidney disease

Kidney disease

Chronic kidney disease affects from 3% to 17% of the population across European Union. 1,2 Regarding health expenditure, for example, Germany is facing more than €3 billion of costs for patients with chronic kidney failure. Therefore, it requires a very costly therapy for the individuals.

Up to 17 %

population affected

> €3 billion

public cost in Germany

Costly therapy for people

Led by Charité, this pilot uses the results of the ongoing MACSS project to improve outcomes and reduce costs after kidney transplantation. Thus intervention, driven by novel dynamic prediction models and alert systems will allow early recognition, management, and prevention of post-transplant complications.

Why a big data approach is needed

In order to deal with the variety of large data sources, reliable natural language processing (NLP) and machine learning (ML) methods are required to allow early prediction of complications and thus facilitate early intervention to prevent morbidity and hospitalizations.
Participants
AOK
Leader
Charite
Dfki
Hpi Hasso Plattner Institut
Universitatskilinikum essen

Publications

Pilot 2: Kidney disease

04/10/2019

Poster at BigMedilytics event: “Big Data: Fueling the transformation of Europe’s Healthcare Sector”. September 4-5, 2019, Valencia, Spain

Exploring Diachronic Changes of Biomedical Knowledge using Distributed Concept Representations

31/08/2019

Vashisth G, Voigt-Antons J-N, Mikhailov M, Roller R. Exploring Diachronic Changes of Biomedical Knowledge using Distributed Concept Representations. Proceedings of the 18th BioNLP Workshop and Shared Task, 2019, Page(s) 348-358. DOI: 10.18653/v1/w19-5037

Knowledge Distillation from Machine Learning Models for Prediction of Hemodialysis Outcome

30/06/2019

Freitas da Cruz, H., Horschig, S., Nusshag, C., Schapranow, M.-P. (2019). Knowledge Distillation from Machine Learning Models for Prediction of Hemodialysis Outcomes. International Journal On Advances in Life Sciences, 11 (12), 33-43.

External Validation of a “Black-Box” Clinical Predictive Model in Nephrology: Can Interpretability Methods Help Illuminate Performance Differences?

30/05/2019

da Cruz H.F., Pfahringer B., Schneider F., Meyer A., Schapranow MP. (2019) External Validation of a “Black-Box” Clinical Predictive Model in Nephrology: Can Interpretability Methods Help Illuminate Performance Differences? In: Riaño D., Wilk S., ten Teije A. (eds) Artificial Intelligence in Medicine. AIME 2019. Lecture Notes in Computer Science, vol 11526. Springer, Cham


1 Source: The global issue of kidney disease. Lancet. 2013 Jul 13;382(9887):101. doi: 10.1016/S0140-6736(13)61545-7.

2 Source: CKD Prevalence Varies across the European General Population. Brück K et al. J Am Soc Nephrol. 2016 Jul;27(7):2135-47.