Project

BigMedilytics: Big Data for Medical Analytics

Goal: BigMedilytics aims to enhance patient outcomes and increase productivity in the health sector by applying big data technologies to complex datasets while ensuring security and privacy of personal data.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 780495.

www.bigmedilytics.eu

Date: 1 January 2018

Updates
0 new
0
Recommendations
0 new
0
Followers
0 new
12
Reads
0 new
89

Project log

Cristina López
added a research item
Aims: This study assessed the impact of acute hemoglobin (Hb) falls in heart failure (HF) patients. Methods: HF patients with repeated Hb values over time were included. Falls in Hb greater than 30% were considered to represent an acute episode of anemia and the risk of hospitalization and all-cause mortality after the first episode was assessed. Results: In total, 45,437 HF patients (54.9% female, mean age 74.3 years) during a follow-up average of 2.9 years were analyzed. A total of 2892 (6.4%) patients had one episode of Hb falls, 139 (0.3%) had more than one episode, and 342 (0.8%) had concomitant acute kidney injury (AKI). Acute heart failure occurred in 4673 (10.3%) patients, representing 3.6/100 HF patients/year. The risk of hospitalization increased with one episode (Hazard Ratio = 1.30, 95% confidence interval (CI) 1.19–1.43), two or more episodes (HR = 1.59, 95% CI 1.14–2.23, and concurrent AKI (HR = 1.61, 95% CI 1.27–2.03). A total of 10,490 patients have died, representing 8.1/100 HF patients/year. The risk of mortality was HR = 2.20 (95% CI 2.06–2.35) for one episode, HR = 3.14 (95% CI 2.48–3.97) for two or more episodes, and HR = 3.20 (95% CI 2.73–3.75) with AKI. In the two or more episodes and AKI groups, Hb levels at the baseline were significantly lower (10.2–11.4 g/dL) than in the no episodes group (12.8 g/dL), and a higher and significant mortality in these subgroups was observed. Conclusions: Hb falls in heart failure patients identified those with a worse prognosis requiring a more careful evaluation and follow-up.
Roland Roller
added a research item
Many people share information in social media or forums, like food they eat, sports activities they do or events which have been visited. This also applies to information about a person's health status. Information we share online unveils directly or indirectly information about our lifestyle and health situation and thus provides a valuable data resource. If we can make advantage of that data, applications can be created that enable e.g. the detection of possible risk factors of diseases or adverse drug reactions of medications. However, as most people are not medical experts, language used might be more descriptive rather than the precise medical expression as medics do. To detect and use those relevant information, laymen language has to be translated and/or linked to the corresponding medical concept. This work presents baseline data sources in order to address this challenge for German. We introduce a new data set which annotates medical laymen and technical expressions in a patient forum, along with a set of medical synonyms and definitions, and present first baseline results on the data.
Roland Roller
added a research item
In research best practices can change over time as new discoveries are made and novel methods are implemented. Scientific publications reporting about the latest facts and current state-of-the-art can be possibly outdated after some years or even proved to be false. A publication usually sheds light only on the knowledge of the period it has been published. Thus, the aspect of time can play an essential role in the reliability of the presented information. In Natural Language Processing many methods focus on information extraction from text, such as detecting entities and their relationship to each other. Those methods mostly focus on the facts presented in the text itself and not on the aspects of knowledge which changes over time. This work instead examines the evolution in biomedical knowledge over time using scientific literature in terms of diachronic change. Mainly the usage of temporal and distributional concept representations are explored and evaluated by a proof-of-concept.
Cristina López
added a research item
Aims: The objective of the present study is to assess the prognostic value of acute kidney injury (AKI) in the evolution of patients with heart failure (HF) using real-world data. Methods and results: Patients with a diagnosis of HF and with serial measurements of renal function collected throughout the study period were included. Estimated glomerular filtration rate (GFR) was calculated with the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). AKI was defined when a sudden drop in creatinine with posterior recovery was recorded. According to the Risk, Injury, Failure, Loss, and End-Stage Renal Disease (RIFLE) scale, AKI severity was graded in three categories: risk [1.5-fold increase in serum creatinine (sCr)], injury (2.0-fold increase in sCr), and failure (3.0-fold increase in sCr or sCr > 4.0 mg/dL). AKI incidence and risk of hospitalization and mortality after the first episode were calculated by adjusting for potential confounders. A total of 30 529 patients with HF were included. During an average follow-up of 3.2 years, 5294 AKI episodes in 3970 patients (13.0%) and incidence of 3.3/100 HF patients/year were recorded. One episode was observed in 3161 (10.4%), two in 537 (1.8%), and three or more in 272 (0.9%). They were more frequent in women with diabetes and hypertension. The incidence increases across the GFR levels (Stages 1 to 4: risk 7.6%, 6.8%, 11.3%, and 12.5%; injury 2.1%, 2.0%, 3.3%, and 5.5%; and failure 0.9%, 0.6%. 1.4%, and 8.0%). A total of 3817 patients with acute HF admission were recorded during the follow-up, with incidence of 38.4/100 HF patients/year, 3101 (81.2%) patients without AKI, 545 (14.3%) patients with one episode, and 171 (4.5%) patients with two or more. The number of AKI episodes [one hazard ratio (HR) 1.05 (0.98-1.13); two or more HR 2.01 (1.79-2.25)] and severity [risk HR 1.05 (0.97-1.04); injury HR 1.41 (1.24-1.60); and failure HR 1.90 (1.64-2.20)] increases the risk of hospitalization. A total of 10 560 deaths were recorded, with incidence of 9.3/100 HF patients/year, 8951 (33.7%) of subjects without AKI episodes, 1180 (11.17%) of subjects with one episode, and 429 (4.06%) with two or more episodes. The number of episodes [one HR 1.05 (0.98-1.13); two or more HR 2.01 (1.79-2.25)] and severity [risk 1.05 confidence interval (CI) (0.97-1.14), injury 1.41 (CI 1.24-1.60), and failure 1.90 (CI 1.64-2.20)] increases mortality risk. Conclusions: The study demonstrated the worse prognostic value of sudden renal function decline in HF patients and pointed to those with more future risk who require review of treatment and closer follow-up.
Alexandra Muñoz
added a project goal
BigMedilytics aims to enhance patient outcomes and increase productivity in the health sector by applying big data technologies to complex datasets while ensuring security and privacy of personal data.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 780495.