Automated Segmentation of Peritoneal Cancer Index Areas Using Deep Learning from CT Scans
Clinical Scorecard: Automated Segmentation of Peritoneal Cancer Index Areas Using Deep Learning from CT Scans
At a Glance
Category Detail
Condition Peritoneal Metastases Key Mechanisms Deep learning for automated segmentation of rPCI regions from CT scans. Target Population Patients diagnosed with gastric, colorectal, or ovarian cancer. Care Setting Clinical imaging and assessment of peritoneal cancer.
Key Highlights
Introduction of a deep learning method for segmenting rPCI regions. Comparison of convolutional and transformer-based architectures. Development of an anatomically constrained pipeline for segmentation. Evaluation of performance against interobserver agreement. Guideline-Based Recommendations
Diagnosis
Use diagnostic laparoscopy as the gold standard for PM diagnosis. Apply the Sugarbaker’s Peritoneal Cancer Index (PCI) for grading PM severity. Management
Consider cytoreductive surgery combined with hyperthermic intraperitoneal chemotherapy. Monitoring & Follow-up
Utilize imaging-based evaluations to assess PM presence and extent. Risks
Small lesions may be missed in imaging evaluations. Patient & Prescribing Data
Patients with peritoneal metastases from gastric, colorectal, or ovarian cancers.
Noninvasive methods for evaluating PM are increasingly needed.
Clinical Best Practices
Ensure precise assessment of PM extent for treatment eligibility. Adopt standardized radiological evaluations based on rPCI definitions. Related Resources & Content
