Detection of Gallbladder Disease Types Using Deep Learning: An Informative Medical Method.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computer-Aided Diagnosis

Background:

• Early and accurate disease diagnosis is crucial in healthcare.
• Gallbladder diseases pose diagnostic challenges, necessitating advanced tools.
• Deep learning offers potential for analyzing medical images like ultrasound.

Purpose of the Study:

• To develop and evaluate deep learning models for the simultaneous classification of nine gallbladder diseases using ultrasound images.
• To create a comprehensive dataset of gallbladder ultrasound images for training and validation.

Main Methods:

• A balanced dataset of 10,692 gallbladder ultrasound images from 1782 patients was curated from three hospitals.
• Images underwent preprocessing and enhancement for segmentation.
• Four deep neural network models were applied and compared for disease classification.

Main Results:

• All four deep neural network models demonstrated effective gallbladder disease detection.
• The MobileNet model achieved the highest accuracy at 98.35% in classifying nine distinct gallbladder diseases.

Conclusions:

• Deep learning models, particularly MobileNet, show high efficacy in diagnosing multiple gallbladder diseases from ultrasound images.
• This approach facilitates early and accurate detection, potentially improving patient outcomes.