Deep convolution neural network for screening carotid calcification in dental panoramic radiographs

Abstract Ischemic stroke, a leading global cause of death and disability, is caused by carotid arteries atherosclerosis. Such calcifications are classically detected by ultrasound screening. In recent years it was shown that these calcifications can also be inferred from routine panoramic dental radiographs. In this work, we focused on the panoramic dental radiographs taken from 500 patients, manually labelling each of the patients’ sides (each radiograph was treated as two sides), and which were used to develop an artificial intelligence (AI)-based algorithm to automatically detect carotid calcifications. The algorithm uses deep learning convolutional neural networks (CNN), with transfer learning (TL) approaches followed by eXtreme Gradient Boosting algorithm (XGBoost) that achieved true labels for each corner, and reaches a sensitivity (recall) of 0.82 and a specificity of 0.93 for individual artery, and a recall of 0.88 and specificity of 0.86 for individual patients. Applying and integrating the algorithm we developed in healthcare units and dental clinics has the potential of reducing stroke events and their mortality and morbidity consequences. Author summary Stroke is a leading global cause of death and disability. One major cause of stroke is carotid artery calcification (CAC). Traditional approaches for CAC detection are doppler ultrasound screening and angiography computerized tomography (CT), medical procedures that require financial expenses, are time consuming and discomforting to the patient. Of note, angiography CT involves the injection of contrast material and exposure to x-ray ionizing irradiation. In recent years researchers have shown that CAC can also be detected when analyzing routine panoramic dental radiographs, a non-invasive, cheap and easily accessible procedure. This study takes us one step further, in developing artificial intelligence (AI)-based algorithms trained to detect such calcifications in panoramic dental radiographs. The models developed are based on deep learning convolutional neural networks, transfer learning, and XGBoost algorithm, that enable an accurate automated detection of carotid calcifications, with a recall of 0.82 and a specificity of 0.93. Statistical approaches for assessing predictions per individual ( i . e .: predicting the risk of calcification in at least one artery), were developed showing a recall of 0.88 and specificity of 0.86. Applying and integrating this approach in healthcare units may significantly contribute to identifying at-risk patients.