Deep Learning System Outperforms Surgeons in Identifying Peritoneal Surface Metastases
Development and testing of a prototype deep learning surgical guidance system showed improved performance in identifying peritoneal surface metastases compared to oncologic surgeons. The system demonstrated potential for increasing identification of metastases by 5% and reducing unnecessary biopsies by 28% in a simulated clinical environment.
In patients with gastric cancer and positive peritoneal lavage cytology or localized peritoneal metastasis, upfront chemotherapy followed by surgery yields better outcomes than surgery followed by chemotherapy.
Secondary cytoreductive surgery with hyperthermic intraperitoneal chemotherapy improves survival for select patients with recurrent colorectal cancer peritoneal metastases.