Trend Report Big Data: How digitalization is changing medicine

Blood values, medication plans or brain scans – every day millions of pieces of patient information end up in computers at doctors’ practices and clinics. Since digitalization leads to gigantic amounts of data, it is important not only to collect it, but also to make optimum use of it. In this way researchers, hospital management, medical business and technology services, and ultimately patients, can derive optimum benefit from this information base.

Immer mehr Daten über unsere Gesundheit werden digital gesammelt.
Rising amounts of data about our health are stored digitally. (Credits: ktsdesign /

Wearables, smartwatches and blood pressure apps in mobile phones: These technical medical helpers provide almost all relevant vital data from their owners. On request, the devices record cholesterol and blood pressure rates, measure the pulse or analyze respiratory rates and sleeping phases. All of this data can help a doctor diagnose a disease, but it can also make the patient more transparent. Still, “Big Data” is revolutionizing medicine, which will also change with it. These mountains of data should soon help patients receive more targeted treatments. Behind the “Big Data” slogan is hard work and artificial intelligence – to discover data patterns that have not been noticed so far or that the human brain simply cannot recognize. This can mean enormous potential for the medical field.

Currently, most of the patient data collected in medical facilities is still unused and left in raw form, since any one human can’t do much with these mountains of data. For computers, however, such unstructured data is a coveted raw material that can be easily sorted, combined and analyzed. If sufficient data is available, some computers can detect diseases even before they occur. To achieve this, the computers analyze and evaluate patterns and relationships in large patient data sets.

Supercomputers could help prevent diseases

Doctors hope that it will be theoretically possible to discover the causes of a disease by comparing extensive data from thousands of patients. To do this, however, the data must first be made accessible, and then powerful computers could scour the vast amounts of patient files, specialist articles, protocols and the like, and thus quickly identify correlations, help in the prevention and early detection of diseases. They could even suggest evidence-based diagnoses and therapies. With the help of these computers, patients might be treated faster, while unnecessary treatments and examinations would be eliminated. Clinics, doctors and the entire healthcare system could save costs.

But wherever data is stored or worked with, there are also risks. Doctors warn against relying solely on a computer evaluation. This is firstly because a patient is not just a statistical probability, but a person who deserves a human judgement. On the other hand – unlike clinical studies – computer programs do not directly test hypotheses about cause-effect relationships, but merely reveal correlations between data sets. In order to prove the efficacy and benefits of a treatment, causal relationships are needed.

Smarte Geräte werden immer beliebter. Mit vielen lässt sich auch die eigene Gesundheit tracken.
Smart devices are gaining popularity. Most of them also allow users to track their health.

Potential security risks to privacy

In addition to many uncertainties, data security and data protection are also important, of course, for the collection of large amounts of data. Anyone who collects data must ensure that it is not misused. The data must therefore neither be used commercially nor passed on without permission. Thus, a scenario in which a supercomputer assists a doctor and helps him to link the patient’s symptoms and measured values with new studies and therapy recommendations is still utopian. So far, the IBM supercomputer Watson seems unable to perform any healing wonders in the treatment of tumors, for example. In 2013, a major project was announced for a Watson-based oncological expert system at the MD Anderson Cancer Center at the University of Texas, but the project has since been abandoned. It is possible that the goals for AI in medicine were initially set too high.

Millions of patient data sets for a genome project

To some extent, the use of Big Data in medicine is already a reality today. Examples include the large genome collections that are already used in systems medicine. The British “100,000 Genomes Project” aims to decode 100,000 genomes. The Netherlands has announced the “GoNL” project (Genome of the Netherlands), Saudi Arabia has created the “Saudi Human Genome Program” and, in the United States, former president Barack Obama announced the “Precision Medicine Initiative” in 2015 with the aim of collecting genetic and medical data of one million people. The genome sequencing forms a data pool whose use will enable new scientific discoveries and medical knowledge for the future.

Another example of Big Data in use is found in pharmaceutical research. Normally, the development of a new drug takes years or decades. However, the use of computers and intelligent algorithms can accelerate pharmaceutical and medical research in almost every phase of drug development, from the idea to clinical trials. As an example, a team from the Mittelhessen research campus hopes to use data from patient registers to identify existing gaps in care for Crohn’s disease and ulcerative colitis, and to help improve medical care for children and young people. Under the leadership of Dr. Jan De Laffolie, Head of Children’s Gastroenterology at the University Hospital of Giessen and Marburg (UKGM), the existing data will be consolidated and supplemented by new data sets and analysis methods in order to find out whether individual care deficits can be reduced.

Prof. Dr. Keywan Sohrabi and Prof. Dr. Volker Gross from the Department of Health at the Technische Hochschule Mittelhessen (THM) also are also contributing their expertise in the fields of Big Data, data mining and E-Health to the Mittlehessen research network. “Today, medical informatics is able to process huge amounts of data. This opens up completely new perspectives for the future of medicine. For example, we can draw conclusions for individual therapy from the analysis of anonymous patient databases,” explains Prof. Sohrabi. The team of scientists is supported by Prof. Dr. Henning Schneider, Dean of the Department of Health at the THM and Director of the Institute for Medical Informatics at the Justus Liebig University in Giessen (JLU). Only in this way can research, hospital management, industry and – ultimately – patients optimally benefit from it.

Leave a Comment