This paper describes the development of a computer-aided diagnostic (CAD) system for assessing uterine conditions. The MATLAB environment was used to calculate features for image and feature extraction. While some feature functions were readily available, most had to be manually developed using mathematical formulas and algorithms to compute specific features. The process begins with pre-processing ultrasound images, which involves resizing them to 200 x 200 pixels. Pre-existing functions are then applied to extract relevant features. After several experimental iterations to optimize results, principal component analysis (PCA) was employed to reduce 40,000 features to 198, which were then used for training. The proposed method classifies images using artificial neural networks (ANN) and provides an interface that demonstrates the system’s functionality by inputting feature information. The main steps include image extraction, comparison with matrix values, and classification into categories like normal, fibroid, or ovarian cyst. The significance of this research lies in its potential to assist radiologists and doctors in accurately identifying the location and type of uterine diseases, thereby reducing the risk of misdiagnoses that could jeopardize patient health.