Applying Artificial Intelligence to Optimize Early-stage Hepatocellular Carcinoma Treatment Based on Multi-modal Image

Artificial Intelligence to Optimize Early-stage Hepatocellular CarcinomaTreatment Based on Multi-modal Image

This study intends to establish two prognostic models based on contrast-enhanced ultrasound (CEUS) and dynamic enhanced magnetic resonance (DE-MRI) multimodal images: prognostic model of liver cancer patients after hepatectomy and prognostic model of liver cancer patients after radiofrequency ablation. Combined with artificial intelligence imaging omics, traditional imaging omics and clinical information, to predict and compare the prognosis of two different treatment methods for early liver cancer, so as to realize the individual selection of treatment methods for early liver cancer patients

Inclusion Criteria:

• Primary single hepatocellular carcinoma confirmed by histology or cytology, with a maximum diameter ≤5.0 cm;

  • Good liver function, Child-Pugh grade A;

    • No serious dysfunction of heart, lung, kidney and other important organs ④ Liver resection or radiofrequency ablation was performed in our hospital, and the study protocol and follow-up procedure were followed.

Exclusion Criteria:

• Patients who cannot perform DCE-MRI examination due to metal implants in the body ② Patients who had received other treatment before DCE-MRI or CEUS;

  • Invisible lesions, diffuse lesions or poor DCE-MRI/CEUS image quality under CEUS; ④ Drug abuse, clinical or psychological or social factors that make informed consent or study implementation affected.