Development and Validation of a PSN-AI Prediction Model

Development and Validation of a AI Prognostic Model Based on DNA Ploidy, Stroma, and Nucleotyping for Synchronous Peritoneal Metastasis From Colorectal Cancer

Heterogeneity concerning survival in synchronous peritoneal metastasis from patients with colorectal cancer exists, thereby further classification is urgently required. This study aimed to validate a PSN-AI model based on DNA ploidy, stroma-tumor fraction, and nucleotyping (PSN) for the prediction of survival of synchronous peritoneal metastasis from colorectal cancer.

Inclusion Criteria:

• Histological proof of synchronous peritoneal metastases of a nonappendiceal colorectal adenocarcinoma.
• Resectable disease determined by computed tomography (CT) and a peritoneal cancer index (PCI) of ≤20 at diagnostic laparoscopy or laparotomy;
• No evidence of systemic (e.g. liver, lung) colorectal metastases within three months prior to enrolment;
• Undergoing cytoreduction surgery for synchronous peritoneal metastasis from colorectal cancer.

Exclusion Criteria:

• Combined with other types of malignant tumors.