Validating New Software for Semiautomated Liver Volumetry--Better than Manual Measurement?

Area of Science:

• Medical Imaging
• Surgical Planning
• Quantitative Analysis

Background:

• Accurate liver volumetry is crucial for surgical planning, especially in cases requiring partial liver resection (hemihepatectomy).
• Traditional manual methods for liver volume measurement can be time-consuming and operator-dependent.
• The development of automated or semiautomated tools aims to enhance the efficiency and reproducibility of these measurements.

Purpose of the Study:

• To prospectively compare the performance of a manual liver volumetry program against semiautomated software.
• To evaluate differences in speed, accuracy, and user experience dependency between the two methods.
• To determine if semiautomated software offers advantages for preoperative liver volume prediction.

Main Methods:

• A prospective study involving ten patients undergoing hemihepatectomy, with IRB approval and informed consent.
• Preoperative 3-phase CT scans were used for whole liver volumetry and resection volume prediction.
• Liver volumes were measured using both a manual (slice-by-slice user definition) and a semiautomated (automatic identification with manual assistance) software approach.
• Six observers with varying experience levels performed the measurements.
• Water displacement volumetry of the resected liver served as the gold standard.

Main Results:

• Both manual and semiautomated methods showed excellent correlation with water displacement volumetry (ρ = 0.997 and ρ = 0.995, respectively).
• Semiautomated software achieved significantly smaller differences between predicted and actual resected volumes (33% vs. 57%, p = 0.002).
• The semiautomated software was nearly four times faster for whole liver volumetry (1:47 min vs. 6:59 min).

Conclusions:

• Both manual and semiautomated software provide accurate estimations of liver volume.
• Semiautomated software demonstrates superior accuracy in predicting the volume of the resected liver portion.
• The semiautomated approach is faster, more reproducible, and less dependent on the operator's experience, making it advantageous for clinical application.