Diagnostic Performance of Deep Learning Applications in Hepatocellular Carcinoma Detection Using Computed Tomography...

Area of Science:

• Medical Imaging
• Artificial Intelligence in Oncology
• Hepatocellular Carcinoma Research

Background:

• Hepatocellular carcinoma (HCC) is a leading cause of cancer mortality globally.
• Late diagnosis is a primary factor contributing to poor patient outcomes in HCC.
• Early detection of HCC remains a significant clinical challenge.

Purpose of the Study:

• To evaluate the efficacy of a deep learning (DL) model, specifically the You Only Look Once (YOLO) architecture, for detecting HCC in computed tomography (CT) images.
• To enhance early diagnosis of HCC through advanced AI techniques.
• To improve patient outcomes by reducing diagnostic delays and errors.

Main Methods:

• A dataset comprising 1290 CT images from 122 patients was utilized.
• The dataset was split into training (70%), validation (20%), and testing (10%) sets.
• A YOLO-based DL model was trained and validated for HCC detection in CT scans.

Main Results:

• The YOLO-based DL model achieved high diagnostic accuracy: precision of 0.972, recall of 0.919, and overall accuracy of 95.35%.
• The model demonstrated excellent performance with a specificity of 95.83% and a sensitivity of 94.74%.
• These results significantly surpass traditional diagnostic methods in accuracy and efficiency.

Conclusions:

• The YOLO architecture shows substantial promise for the early detection of HCC in CT imaging.
• DL models integrated with AI technology are poised to become standard tools in oncological diagnostics.
• Advancements in AI diagnostics can improve accuracy, efficiency, and patient survival rates in oncology.