Bones are both strong and flexible tissue made up of collagen and calcium phosphate. They mainly contribute to the movement of the human body and also serve as a protective shield for the body's soft organs such as the heart, lungs, and brain. Without them, the human body would not be built to function normally. But sometimes, due to various accidents, a person is exposed to some diseases such as infection or injury, which lead to defects in normal growth and shape of bone structure. This defect in the bone structure is called bone abnormalities, which increase the risk of fractures, cause chronic pain or disability, and in some cases, the risk of death if not diagnosed in a timely, accurate and quick manner so that the specialists make the right decision to treat the patient. Often, the initial diagnosis of bone abnormalities is made by doctors and specialists using X-rays of the patient's complaint and the injury site to show the X-ray of the shape and density of the bones, then they are classified into normal or abnormal. Detection and classification depend on experience and human effort, so the error in the results of this process can expose the patient to a great danger and catastrophe of his life. Therefore, deep learning algorithms from artificial intelligence and machine learning science were applied to help specialists avoid wrong or inaccurate diagnoses when detecting bone abnormalities in X-ray images by using a pre-trained convolutional neural network called VGG16. The model was customized to fit the bone abnormalities classification then applied to a dataset consisting of 42000 X-rays of the upper bones of some patients collected from Kaggle depository called Mura-v1.1. We trained, validated, and tested the modified VGG16 model. The proposed VGG16 mode obtained Precision (85.96%), Recall (85.82%) and F1-Score (85.77%).