A new deep learning based data pipeline and cloud architecture for cervical spine fracture detection.

Abstract

Cervical spine fractures are a serious medical emergency that can lead to permanent paralysis or even death. Moreover, rapid and accurate detection of such fractures is essential for optimal patient care. However, manually interpreting Computed Tomography (CT) scans for detecting possible fractures in the cervical spine, as done traditionally, is time-consuming and requires expertise from experienced radiologists. In this context, Artificial Intelligence has the potential to revolutionize cervical spine fracture detection by providing fast, accurate, and automated solutions. In this manuscript, we have studied a couple of contributions. In the first contribution, we have performed a review of the essential works established in the literature in the setting of the addressed issue. Specifically, we have analyzed, discussed, and compared them according to appropriate criteria. In the second contribution, we have developed a multifaceted computational pipeline based on the combination of Faster R-CNN and NeXt-ViT models in view of detecting fractures within the cervical spine. We have trained and evaluated the proposed pipeline on the large RSNA public dataset containing cervical spine CT scans. Hence, the new system has achieved encouraging results. Furthermore, the new proposed data pipeline’s ability to detect subtle and complex fractures has motivated us to integrate it in a cloud-based architecture that we present in the framework of this work. The proposed cloud-based architecture has the potential to be used as a distant clinical decision-support tool to help radiologists identify fractures quickly and reliably, and to be continuously improved through a feedback mechanism.