Big Data analytics, the problems and the AEGLE solution

Currently, there is a lot of interest in big data analytics as a solution for many problems in healthcare. Both industry and academia have invested much these past years to make it all possible. However, many barriers remain before the benefits from big data can be reaped such as data availability, integration, analysis, and wider acceptance of big-data solutions. AEGLE aims to develop a big data analytics platform that addresses almost the entire big data value chain including data analysis, curation, storage and usage (Curry et al. 2016).

Big data applications

Oncology (Halamka et al. 2014), genetic research (Phillips et al. 2014) and acute care settings (Bates et al. 2014), are clinical cases where the adoption of big-data is expected to bring serious impact. Large companies such as IBM have invested in big data analysis in cancer care resulting in interesting innovations such as Watson for Healthcare. But most of the services offered are still in the early pilot phase while adopting cloud platforms for big data analysis is not yet standard procedure. Many initiatives are public and more recent such as BigData@Heart (BigData@Heart 2018) and BigO (BigO 2018). Combined, the initiatives focus on a wide variety of users such as clinical practitioners, industry researchers, and policymakers in a variety of clinical settings from oncology to acute care and non-malignant chronic diseases.

The AEGLE platform

The AEGLE platform addresses challenges related to the volume and complexity of the data (i.e. prolonged duration of analyses) but also challenges faced with privacy and deployment efficiency. With the AEGLE platform, users can pseudo-anonymize their data conforming to EU regulations prior to uploading it onto a cloud-based platform. The platform includes a scalable data management solution offering the users the option of paying only for the storage they use. Moreover, the platform ensures a secure environment for sharing data between colleagues who are located at different sites and countries. The platform focusses on three problems currently identified in the healthcare setting; (1) the inability to interpret excessive amounts of monitoring data in clinical practice, (2) the challenge of handling large amounts of genetic’s data for research and (3) the challenge of interpreting many patient-related parameters in attempts to personalize treatment. Analytics on the platform aims to address these three problems.

First, features are available for analysing NGS data using chronic lymphocytic leukaemia as an indicative case. The platform offers a wide range of analytic pipelines for variant arrangement, calling, annotation and visualization of NGS data such as whole exome and RNA-sequencing data utilising Maxeler’s acceleration technologies. Moreover, a key feature of the platform is the analysis of immunogenetics data.

The platform is also able to analyse monitoring data in an intensive care unit. In current practice, the continuous data stream generated by monitors of vital signs and mechanical ventilators is of limited use to clinicians since real-time, complex analyses are often not possible. The AEGLE platform addresses that by offering a clinical-decision support user interface which provides results from real-time, complex analyses. The aim is that the analytics enable identification of ineffective efforts of mechanically ventilated patients, catheter-related bloodstream infection, suboptimal nutrition adherence and deterioration of the patient. Furthermore, the platform for the intensive care also enables clinical researchers to study their patient population and develop their own predictive analytics.

The final clinical setting, that AEGLE addresses, concerns with chronic diseases. In the electronic health record, data is constantly collected but not exploited for offering personalized care. The analytics provided for non-malignant chronic diseases are developed using data from patients with Diabetes Mellitus Type II. Researchers are given the ability to improve and develop predictive analytics for personalizing care but also to generate heatmaps and classification pyramids for specific patient populations in a hospital.

But what does the customer want?

How is such a platform offered to customers? Some users wish to use the platform complementary to their existing solutions. Meanwhile, others will use it to replace the software solutions they currently have in practice. Furthermore, what would a user be willing to invest in such a new solution?

AEGLE offers you as a potential user to share your thoughts on how the platform should be offered to customers.
Provide us with your opinion of the AEGLE solution by filling in THIS questionnaire! It will only take 5-10 minutes.

Learning from other initiatives; many solutions to solve a problem

The AEGLE platform addresses several challenges but there are still many issues to overcome before benefits of big data analytics can be maximized. It is therefore important for those that aim to develop big data analytics to collaborate and learn from other solutions. Most of the big data initiatives and commercial entities address challenges of the big data value chain, such as the aggregation of data and real-time visualization of risk factors, in their own unique way.

These initiatives illustrate that there are many solutions to solve a single problem. To stimulate such initiatives to collaborate and present their work, the AEGLE project will host a small conference where initiatives have the opportunity to provide demonstrations of their prototypes and present their ideas.

If you participate in a big data project, start-up or initiative; do not hesitate to submit an abstract HERE!

References

Bates DW, Saria S, Ohno-Machado L, Shah A, Escobar G. Big data in healthcare: Using analytics to identify and manage high-risk and high-cost patients. Health Aff (Millwood) 2014 Jul;33(7):1123-31.

BigData@Heart [Internet] Netherlands: UMC Utrecht; c2018 [cited 2018. Available from: https://www.bigdata-heart.eu/

BigO: Big Data Against Childhood Obesity [Internet] [cited 2018. Available from: https://bigoprogram.eu/.

Curry E. The big data value chain: Definitions, concepts, and theoretical approaches. In: New horizons for a data-driven economy. Springer; 2016.

Halamka JD. Early experiences with big data at an academic medical center. Health Aff (Millwood) 2014 Jul;33(7):1132-8.

Phillips KA, Trosman JR, Kelley RK, Pletcher MJ, Douglas MP, Weldon CB. Genomic sequencing: Assessing the health care system, policy, and big-data implications. Health Aff (Millwood) 2014 Jul;33(7):1246-53.

Author:
Lytske Bakker
Erasmus School of Health Policy & Management (Erasmus University Rotterdam)