Fraunhofer Austria AI technique cuts liver cancer surgery time

Researchers at Fraunhofer Austria have combined optical coherence tomography (OCT) with artificial‑intelligence‑based anomaly detection to analyze liver tissue during cancer operations. In a study of 173 scans from 69 patients, the system distinguished healthy from tumor tissue in seconds, achieving up to 94 % accuracy for certain cancer types and an overall correct classification rate of 81 %.

The rapid intra‑operative diagnosis could shorten anesthesia duration, reduce staff workload and lower complication risk. Dr. Ulrich Krispel explained, “Das Besondere an dieser Methode ist, dass für das Training dieser KI nur Gutdaten … die Methode erkennt dann Abweichungen von der Norm sehr zuverlässig.” PD Dr. Iakovos Amygdalos added, “Ich denke dieser Ansatz bietet großes Potenzial … die Operation deutlich verkürzen, das Personal entlasten und den Eingriff patientenfreundlicher machen.” The findings were published in Scientific Reports and represent the first use of AI‑driven anomaly detection on OCT images of liver tissue.