A new software tool for planning interventional procedures in liver cancer

Area of Science:

• Medical Imaging
• Interventional Radiology
• Computational Anatomy

Background:

• Intra-arterial therapies (chemotherapy, radiation) are effective for primary liver cancer (hepatocellular carcinoma).
• Minimally invasive intra-arterial approaches require precise pre-operative planning and intra-operative guidance.
• Existing planning tools may benefit from enhanced segmentation and visualization capabilities for liver cancer interventions.

Purpose of the Study:

• To develop and integrate a semi-automatic image segmentation pipeline for liver cancer treatment planning.
• To create a liver-specific workflow within endovascular planning software (EndoSize®).
• To provide tools for pre-operative planning and intra-operative assistance in liver cancer intra-arterial therapy.

Main Methods:

• Semi-automatic segmentation of main arterial branches from computed tomography angiography (CTA) images.
• Automatic segmentation of the liver, arterial, and venous trees.
• Interactive segmentation of tumors for procedure-specific planning, integrated into EndoSize® software.

Main Results:

• Successful main arterial branch extraction in 85% of 172 patients.
• Quantitative assessment yielded a mean DICE score of 0.91 for liver segmentation.
• Mean DICE scores of 0.47 for venous trees and 0.92 for arterial trees were achieved on public databases.

Conclusions:

• The developed pipeline facilitates direct access to the treatment site for liver cancer therapies.
• It provides crucial anatomic measurements and visualization of hepatic trees, liver, and arteries.
• This approach enhances pre-operative planning for intra-arterial liver cancer treatments.