Mask refinement network for tooth segmentation on panoramic radiographs.

OBJECTIVES: Instance-level tooth segmentation extracts abundant localization and shape information from panoramic radiographs (PRs). The aim of this study was to evaluate the performance of a mask refinement network that extracts precise tooth edges. METHODS: A public dataset which consists of 543 PRs and 16211 labelled teeth was utilized. The structure of a typical Mask Region-based Convolutional Neural Network (Mask RCNN) was used as the baseline. A novel loss function was designed focus on producing accurate mask edges. In addition to our proposed method, 3 existing tooth segmentation methods were also implemented on the dataset for comparative analysis. The average precisions (APs), mean intersection over union (mIoU), and mean Hausdorff distance (mHAU) were exploited to evaluate the performance of the network. RESULTS: A novel mask refinement region-based convolutional neural network was designed based on Mask RCNN architecture to extract refined masks for individual tooth on PRs. A total of 3311 teeth were correctly detected from 3382 tested teeth in 111 PRs. The AP, precision, and recall were 0.686, 0.979, and 0.952, respectively. Moreover, the mIoU and mHAU achieved 0.941 and 9.7, respectively, which are significantly better than the other existing segmentation methods. CONCLUSIONS: This study proposed an efficient deep learning algorithm for accurately extracting the mask of any individual tooth from PRs. Precise tooth masks can provide valuable reference for clinical diagnosis and treatment. This algorithm is a fundamental basis for further automated processing applications.

Ultrasound with microbubbles improves memory, ameliorates pathology and modulates hippocampal proteomic changes in a triple transgenic mouse model of Alzheimer's disease.

Alzheimer's disease (AD) is a progressive neurodegenerative disease manifested by cognitive impairment. As a unique approach to open the blood-brain barrier (BBB) noninvasively and temporarily, a growing number of studies showed that low-intensity focused ultrasound in combination with microbubbles (FUS/MB), in the absence of therapeutic agents, is capable of ameliorating amyloid or tau pathology, concurrent with improving memory deficits of AD animal models. However, the effects of FUS/MB on both the two pathologies simultaneously, as well as the memory behaviors, have not been reported so far. Methods: In this study, female triple transgenic AD (3×Tg-AD) mice at eight months of age with both amyloid-ß (Aß) deposits and tau phosphorylation were treated by repeated FUS/MB in the unilateral hippocampus twice per week for six weeks. The memory behaviors were investigated by the Y maze, the Morris water maze and the step-down passive avoidance test following repeated FUS/MB treatments. Afterwards, the involvement of Aß and tau pathology were assessed by immunohistochemical analysis. Neuronal health and phagocytosis of Aß deposits by microglia in the hippocampus were examined by confocal microscopy. Further, hippocampal proteomic alterations were analyzed by employing two-dimensional fluorescence difference gel electrophoresis (2D-DIGE) combined with mass spectrometry. Results: The three independent memory tasks were indicative of evident learning and memory impairments in eight-month-old 3×Tg-AD mice, which developed intraneuronal Aß, extracellular diffuse Aß deposits and phosphorylated tau in the hippocampus and amygdala. Following repeated FUS/MB treatments, significant improvement in learning and memory ability of the 3×Tg-AD mice was achieved. Amelioration in both Aß deposits and phosphorylated tau in the sonicated hemisphere was induced in FUS/MB-treated 3×Tg-AD mice. Albeit without increase in neuron density, enhancement in axonal neurofilaments emerged from the FUS/MB treatment. Confocal microscopy revealed activated microglia engulfing Aß deposits in the FUS/MB-treated hippocampus. Further, proteomic analysis revealed 20 differentially expressed proteins, associated with glycolysis, neuron projection, mitochondrial pathways, metabolic process and ubiquitin binding etc., in the hippocampus between FUS/MB-treated and sham-treated 3×Tg-AD mice. Conclusions: Our findings reinforce the positive therapeutic effects on AD models with both Aß and tau pathology induced by FUS/MB-mediated BBB opening, further supporting the potential of this treatment regime for clinical applications.