RESUMEN
This study aimed to develop an ultraminiature pressure sensor array to measure the force exerted on teeth. Orthodontic force plays an important role in effective, rapid, and safe tooth movement. However, owing to the lack of an adequate tool to measure the orthodontic force in vivo, it remains challenging to determine the best orthodontic loading in clinical and basic research. In this study, a three-dimensional (3D) orthodontic force detection system based on piezoresistive absolute pressure sensors was designed. The 3D force sensing array was constructed using five pressure sensors on a single chip. The size of the sensor array was only 4.1 × 2.6 mm, which can be placed within the bracket base area. Based on the barometric calibration, conversion formulas for the output voltage and pressure of the five channels were constructed. Subsequently, a 3D linear mechanical simulation model of the voltage and stress distribution was established using 312 tests of the applied force in 13 operating modes. Finally, the output voltage was first converted to pressure and then to the resultant force. The 3D force-detection chip was then tested to verify the accuracy of force measurement on the teeth. Based on the test results, the average output force error was only 0.0025 N (0.7169%) (p = 0.958), and the average spatial positioning error was only 0.058 mm (p = 0.872) on the X-axis and 0.050 mm (p = 0.837) on the Y-axis. The simulation results were highly consistent with the actual force applied (intraclass correlation efficient (ICC): 0.997-1.000; p < 0.001). Furthermore, through in vivo measurements and a finite element analysis, the movement trends generated when the measured orthodontic forces that acted on the teeth were simulated. The results revealed that the device can accurately measure the orthodontic force, representing the first clinical test of an orthodontic-force monitoring system. Our study provides a hardware basis for clinical research on efficient, safe, and optimal orthodontic forces, and has considerable potential for application in monitoring the biomechanics of tooth movement.
RESUMEN
Introduction: Accurate sleep staging is an essential basis for sleep quality assessment and plays an important role in sleep quality research. However, the occupancy of different sleep stages is unbalanced throughout the sleep process, which makes the EEG datasets of different sleep stages have a class imbalance, which will eventually affect the automatic assessment of sleep stages. Method: In this paper, we propose a Residual Dense Block and Deep Convolutional Generative Adversarial Network (RDB-DCGAN) data augmentation model based on the DCGAN and RDB, which takes two-dimensional continuous wavelet time-frequency maps as input, expands the minority class of sleep EEG data and later performs sleep staging by Convolutional Neural Network (CNN). Results and discussion: The results of the CNN classification comparison test with the publicly available dataset Sleep-EDF show that the overall sleep staging accuracy of each stage after data augmentation is improved by 6%, especially the N1 stage, which has low classification accuracy due to less original data, also has a significant improvement of 19%. It is fully verified that data augmentation by improving the DCGAN model can effectively improve the classification problem of the class imbalance sleep dataset.