Deep Neural Network Based Artificial Intelligence Assisted Diagnosis of Bone Scintigraphy for Cancer Bone Metastasis

Abstract Background: Bone scintigraphy (BS) is one of the most frequently utilized diagnostic techniques in detection of cancer bone metastasis, and it occupies great workload for nuclear physicians. So we aim to architecture an automatic image interpreting system to assist physicians for diagnosis.Methods: We developed an artificial intelligence (AI) model based on a deep neural network with 12222 cases of 99mTc-MDP bone scintigraphy, and evaluated its diagnostic performance of bone metastases.Results: This AI model demonstrated diagnostic performance by areas under the curve (AUC) of receiver operating characteristic (ROC) was 0.988 for breast cancer, 0.955 for prostate cancer, 0.957 for lung cancer, and 0.971 for other cancers. Applying this AI model to a new dataset of 400 BS cases, it represented comparable performance to that of human physicians in individually classifying bone metastasis. Further AI-consulting interpretation also improved human diagnostic sensitivity and accuracy.Conclusion: In total, this AI model performed valuable benefit for nuclear physicians in timely and accurate evaluation of cancer bone metastases.