Spleen Segmentation and Assessment in CT Images for Traumatic Abdominal Injuries

Area of Science:

• Medical Imaging
• Radiology
• Computer-Aided Diagnosis

Background:

• Spleen segmentation in computed tomography (CT) images is difficult due to similar gray levels of abdominal organs.
• Manual analysis of CT scans for abdominal trauma is time-consuming, delaying treatment.
• Automating spleen segmentation can significantly reduce diagnosis and treatment times.

Purpose of the Study:

• To develop a fully automated, knowledge-based technique for accurate spleen segmentation in CT images.
• To enable reliable spleen detection in both healthy and injured patients.
• To improve the efficiency of abdominal trauma assessment.

Main Methods:

• A fully automated, knowledge-based approach utilizing anatomical information for spleen segmentation.
• A spleen detection procedure designed for both healthy and injured cases.
• An edge merging technique applied in cases of splenic hemorrhage and laceration.

Main Results:

• The proposed method accurately segments the spleen in CT images.
• Performance was evaluated using metrics including mis-segmented area, accuracy, specificity, and sensitivity.
• The automated method demonstrated superior performance compared to existing techniques.

Conclusions:

• The developed automated spleen segmentation method is effective and accurate.
• This technique shows promise for improving the assessment of abdominal trauma.
• The method offers a significant advancement over current spleen segmentation approaches.