Three examples of optimizations in the hospital with Big Data and AI
A 2020 Deloitte survey found that the COVID-19 pandemic has repeatedly accelerated digital transformation in hospitals and associated facilities.
A third of those at the time indicated that more technology was being used, at least in part. In many places, existing digitization efforts were pushed forward more aggressively, so that many hospitals are now also interested in the added value that comes from the analysis of real-time data. According to a survey by the industry association Bitkom from 2021, 82 percent of doctors in hospitals state that “more speed is needed in the expansion of digital offers”.
Solutions that use AI and data in near real-time enable improvements in a number of use cases, some of which are significant. Three of the most exciting cases are presented here.
Monitoring of diabetic patients
Some doctors are already using analyzes from magnetic resonance imaging (MRI) for the early detection of diabetes. Based on the fat distribution in the body, they can make quick diagnoses and thus give treatment recommendations. Every diabetes patient collects their own data every year. Up to 100,000 glucose measurements are taken per patient per year. If this data were collected centrally and evaluated with AI, doctors could make better and faster decisions about the treatment of the people concerned. In Germany, there are 7.2 percent of the 18 to 79 age group with diabetes, so 2.1 percent are suspected of having diabetes without even knowing it.
Doctors and nurses under stress – does it have to be?
Not only because of the stress of the COVID-19 pandemic, the work intensity and stress of doctors and nurses has increased significantly. The risks of patients suffering from delayed treatments, postponed surgeries and overloaded laboratories are high. With AI and real-time data analytics, faster diagnoses of conditions through comparative data from similar cases could lead to better patient care and subsequent faster recovery. An example of this is dying App imitoWound. It can suggest appropriate treatment to doctors through automatic wound detection and qualitative and quantitative characterization. The project was developed and brought to the Swiss market by the Center for Wounds and Wound Healing at the Geneva School of Health (HEdS-Geneva), the Interprofessional Center for Simulation, the Computer Science Department of the University of Geneva and imito AG.
Smartwatch & wearables – treasure trove of data waiting to be raised
Smartwatches, fitness trackers and other wearables have potential with their countless sensors to send health data to doctors quickly and easily. Analysis skills are still lacking. According to Statista, 3.2 million smartwatches are in use in Germany. They measure heart rate, pulse and some even oxygen saturation, all critical information for doctors who need to save a patient’s life in the event of a stroke. Further development could create many more possibilities here, in which health data through AI and data analyzes in almost real time would give patients faster and more targeted treatment and doctors and nurses would be less stressed by complex and lengthy tests.
Medical advancement based on a Lakehouse architecture
As already written, data and artificial intelligence can change the German healthcare system. However, solid data analysis is only possible when the right data architecture is in place – an open architecture that enables healthcare organizations to manage structured, unstructured and semi-structured data in one place. Dubbed the Data Lakehouse, this data architecture already enables many healthcare organizations around the world to predict disease risk, classify medical images, and accelerate new drug development. Other data architectures, such as data warehouses, are too expensive for the healthcare sector because they impede the delivery of real-time analytical results needed for critical care decisions and are often too complex and costly to manage.
Conclusion
AI and real-time data analysis, supported by a Lakehouse architecture, can immensely improve the possibilities of patient care in Germany. Modern technologies and the data that has been collected over the years – some collected by the patients themselves – not only raise optimization treasures, but also promote the health of the patients themselves. Treatment can be significantly improved through faster diagnoses and less stressed staff.
Author: Michael Sanky, Global Industry Lead, HealthCare & Life Sciences at Databricks, www.databricks.com/de/