A Deep Learning Radiomics Model for Predicting Occult Peritoneal Metastases of Pancreatic Adenocarcinoma

Development and Validation of a Deep Learning Radiomics Model With Clinical-radiological Characteristics for the Identification of Occult Peritoneal Metastases in Patients With Pancreatic Ductal Adenocarcinoma

Occult peritoneal metastases (OPM) in patients with pancreatic ductal adenocarcinoma (PDAC) are frequently overlooked during imaging. We aimed to develop and validate a CT-based deep learning-based radiomics (DLR) model with clinical-radiological characteristics to identify OPM in patients with PDAC before treatment.

This retrospective, bicentric study included 302 patients with PDAC (training: n = 167, OPM-positive, n=22; internal test: n = 72, OPM-positive, n=9: external test, n=63, OPM-positive, n=9) who had undergone baseline CT examinations between January 2012 and October 2022. Handcrafted radiomics (HCR) and DLR features of the tumor and HCR features of peritoneum were extracted from CT images. Mutual information and least absolute shrinkage and selection operator algorithms were used for feature selection. A combined model, which incorporated the selected clinical-radiological, HCR, and DLR features, was developed using a logistic regression classifier using data from the training cohort and validated in the test cohorts.

Inclusion Criteria:

Patients with suspected pancreatic tumors who underwent contrast enhanced CT and pathological examinations at Center 1 and Center 2 were eligible for inclusion in this study.

Exclusion Criteria:

• (a) pathologically diagnosed PDAC by pathology, (b) time intervals between contrast-enhanced CT and pathology less than 2 weeks; (c) history of pancreatic surgery, (d) history of pancreatic malignancy, and (e) poor CT image quality that undermined peritoneal lesion assessment