Object Detection Model Utilizing Deep Learning to Identify Retained Surgical Gauze in the Body on Postoperative Radiography: Phantom Study

PURPOSE Foreign bodies such as a surgical gauze can be retained in the body after surgery and some cases cannot detected by postoperative radiography. The aim of this study was to develop an object detection model capable postsurgical body. used deep learning using abdominal radiographs, phantom performed evaluate ability automatically detect gauze. MATERIALS AND METHODS constructed Single Shot MultiBox Detector (SSD) 300. In total, 268 images were used: 180 training data, 20 validation additional 34 nongauze test data. To performance model, confusion matrix created accuracy sensitivity calculated. RESULT True-positive (TP) rate, true-negative (TN) false-positive (FP) false-negative (FN) rate 0.92, 1.00, 0.00, 0.08, respectively. Accuracy 0.96, 0.92. CONCLUSION could on with high sensitivity.