Universitäre Altersmedizin FELIX PLATTER

Universitäre Altersmedizin FELIX PLATTER – AI facilitates research: Universitäre Altersmedizin FELIX PLATTER efficiently extracts knowledge from its data.

Geriatric medicine is changing: away from pure acute and rehabilitation medicine and towards prevention. Older people want personalised medicine. Geriatric clinics such as the Universitäre Altersmedizin FELIX PLATTER (UAFP) in Basel utilise digital innovations such as AI for this purpose. At the centre of digital development: linking clinical data, accelerating research. Bechtle and the UAFP use Azure OpenAI Services for this. This automatically translates queries from research into precise SQL queries - and makes the wealth of data available to researchers efficiently and securely.

Project details.

Services

Professional Services
Cloud

Industry

Pharmaceutical / Healthcare

Manufacturer

Microsoft

Company size

250 – 1,500 employees

Challenge.

100 per cent natural language for all data queries - without programming knowledge.

Data is a treasure - but only if it can be harvested. The Universitäre Altersmedizin FELIX PLATTER treats 5,000 inpatients every year. All the data collected (patient, research and clinical data) is stored in a comprehensive medical data warehouse. FELIX PLATTER uses this data in anonymised form for research purposes. Previously, the data queries had to be carried out by technical specialists who were familiar with the data model. They translated the queries into SQL. This cost time and resources. But how can this data be made accessible to a wider audience? How can queries be made in natural language without the support of specialists and without complex SQL queries? How can users get quick and precise answers and refine them with further questions if necessary? The wish: Researchers and experts should be able to carry out their data analyses themselves without specialist technical knowledge, but with maximum efficiency. So that new findings find their way into research and healthcare even faster.

We use YouTube to embed video content on our website. This service may collect data on your activity. For more information, please go to the settings page.

Solution.

There is a suitable answer to the question of what such a solution might look like: generative AI. Bechtle developed a local application for this purpose. It links the local data from the FELIX PLATTER medical data warehouse with various Azure OpenAI services. The result is a system that leaves the sensitive data on site and only accesses structures and documentation. Nevertheless, the researchers' questions are transformed into technical queries using a large language model. The AI creates precise queries and presents the results of the mass data analyses clearly in tabular form. It is even possible to visualise the data. But that's not all: Bechtle is continuously developing the FELIX PLATTER AI app, as the possibilities of AI are developing rapidly and dynamically. This means that requirements can be adapted again and again. And the result is a solution that provides precise answers to complex questions in natural language. Quickly, efficiently and from the data in the in-house

As a non-AI specialist, I found the dialogue with Bechtle very exciting. In particular, being able to express my wishes, which Bechtle then anticipated and implemented technically. We were also shown why certain things don't work. It was a really constructive exchange that helped us move forward. I am convinced that the use of AI will make a significant contribution to facilitating the clarification of further research questions. Access to data is becoming increasingly easier. Although we have an enormous wealth of data, it is important to process and make it available in such a way that researchers can use it effectively for their analyses.

Laurent Wagner, Chief Information Officer, Universitäre Altersmedizin FELIX PLATTER

Business benefits.

Simple queries for researchers, who can formulate their analysis requests in natural language - without IT knowledge.
Faster queries from SQL statements for (authorised) non-technical users.
Precise answers from the data sets of the FELIX PLATTER Data Warehouse, which can be further refined quickly.
More research freedom through direct access to all the data required for analyses and studies.
AI explains queries and thus enables a better understanding of the data.
Maximum data security for all sensitive data, which remains completely local - only structural information is processed by the AI.
Promoting innovation through free resources and rapid knowledge transfer - this drives new projects and ideas forward.

Partner.

About our customer Universitäre Altersmedizin FELIX PLATTER.

The Universitäre Altersmedizin FELIX PLATTER combines the specialised fields of acute geriatric medicine, geriatric psychiatry and rehabilitation under the roof of one of the most innovative buildings of its kind. It offers a wide range of services from prevention and early diagnosis through treatment and therapy to aftercare. FELIX PLATTER works on an interdisciplinary basis and in interprofessional teams in order to provide holistic and individualised care for people in old age - i.e. physically and psychologically tailored to the individual patient. Over 5,000 inpatients are treated at FELIX PLATTER every year. Over 3,000 patients make use of outpatient services at FELIX PLATTER every year. felixplatter.ch