|
1000
|
Abstract/Summary
|
-
In times of growing infuence of artifcial intelligence (AI) in industry—including the pharmaceutical industry—it is of utmost interest to check the possibilities and chances and to what extent these new techniques can help to develop and narrow down scientifc questions. The medical-scientifc rationale for this specifc exercise described hereafter is the availability of new pharmaceutical approaches (CD3-directed monoclonal antibody treatment [1], Sanof) which have shown a disease delaying effect on Type 1 Diabetes mellitus (T1DM). The earlier the detection of potential T1DM patients, the better the substantial prolongation of the T1DM manifestation in these patients with T1DM risk [2][3][4]. With the prolongation of the manifestation of T1DM also the T1DM concomitant diseases caused by hyperglycemia can be prolonged e.g. micro- and macrovascular complications. The availability of new drugs prolonging the manifestation of T1DM, the detection of potential T1DM patients brought the screening into the focus since subjects with risk for T1DM mostly beneft from an early detection and therapy. In this context, we wanted to know the capabilities of a LLM (Large Language Model) to extract relevant information from only a few text sources to answer questions concerning the screening for T1DM.
|
