Artificial Intelligence for Intraoperative Neuromonitoring – Preventing Nerve Damage During Thyroid Surgery

Project »Kismo«: Artificial Intelligence in Thyroid Surgery for the Analysis of Intraoperative Neurophysiological Monitoring

Surgery on the thyroid gland carries a significant risk of injury to the nearby vocal cord nerves. If these nerves are damaged, this can have serious consequences for patients. Intraoperative neuromonitoring (IONM) monitors nerve function during surgery and allows surgeons to respond immediately to impending damage by adjusting their surgical strategy. However, this requires surgeons to quickly and correctly interpret the extensive signal data measured and derive the appropriate strategies from it. This is exactly where the »KISMO« project comes in: it is developing an assistance system that reliably supports physicians in this demanding task.

Real-Time Evaluation of Neural Signals and Digitally Assisted Decision-Making

In the »KISMO« project, we are working with Mainz University Medical Center and medical device manufacturer inomed Medizintechnik GmbH to develop methods and software prototypes for AI-supported neuromonitoring. Neural signal data is recorded in real time, processed, and checked for critical changes. Based on these analyses, the system automatically generates suggestions for possible surgical strategies. This enables surgeons to make informed decisions quickly and optimize the procedure.

Ai-Supported Signal Data Analysis and Clinical Decision Support

Clinical case data with neurosignals and surgical procedures are systematically processed and annotated in the »KISMO« project. Based on this, we develop AI models that detect critical signal changes, identify suitable surgical strategies, and predict the further course of the signal. Concepts of Explainable Artificial Intelligence (XAI) play a central role here: they make it possible to understand how the AI arrives at its results. This allows surgeons to review the proposed strategies in real time and immediately make the most appropriate decision for continuing the surgery. The methods and software prototypes developed are intended to significantly reduce the risk of nerve damage during thyroid surgery – and to maintain the best possible quality of life for patients.

Project Partners

  • Inomed Medical Technology GmbH
  • University Medical Center of Johannes Gutenberg University Mainz
Process Overview Project »KISMO«
© freepik / Fraunhofer ITWM
Process Overview Project »KISMO«