Gallstone is a chronic condition that affects people around the globe. Gallstone disease (GSD) is a major challenge on healthcare systems across the globe, and it is the most widespread diseases among people with abdominal discomfort who are admitted to the emergency rooms. The size of the gallbladder organ ultrasound is affected by a multitude of features, including the layer of the gallbladder. As a result, gallbladder scans from various parameters make it a difficult process. In this paper, we have developed a healthcare informatics system to investigate and detect the gallstones present. A detailed comparison of different algorithms inspired by cutting edge models for object detection is used in this paper. These include faster region-based convolutional neural network (Faster-R-CNN, mask regional convolutional neural network (Mask R-CNN), and single shot detector (SSD). The suggested model, which is based on the mask R-CNN, SSD, Faster R-CNN technique, distinguishes stones in the gallbladder by extracting suggestions from the stone area. The Mask R-CNN model was modelled and trained using various backbone networks. We have used the ultrasound images for the experiment obtained from the medical professionals. The ultrasound images gathered are focused on various contexts such as gender, age, and people from urban and rural areas. The results indicate that the Mask R-CNN with backbone network of Resnet-101-FPN combination outperforms in case of object detection.