A Deep Learning Algorithm for Liver Metastasis Detection at Contrast-enhanced Abdominal CT in Patients with Colorectal Cancer: A Comparative Study with Radiologists

Key Points

• The study aims to evaluate a deep learning algorithm’s effectiveness in detecting liver metastasis using contrast-enhanced abdominal CT scans compared to radiologists.
• Developed a deep learning algorithm for detection of liver metastasis.
• Conducted a comparative analysis with radiologists' performance.
• Evaluated detection rates across varying lesion sizes and clinical conditions.
• Deep learning algorithm achieved liver metastasis detection rates comparable to radiologists.
• Performance remained consistent across different lesion sizes and locations.
• Detection capabilities were effective in various clinical conditions.