Abstract:In this paper, a new concept of fully electrified vehicles, based on the principle that the braking energy can be fully recovered via electrical motor and active safety control so as to improve energy efficiency greatly, was proposed. It reveals that this type of vehicles possesses the characteristics of perception for vehicle motion behaviour and the related intelligent detection methods were also introduced. At the same time, the dynamic control structure of the mechanicalelectrical decoupling in the vehicles was also discussed. Furthermore, the strategy of energy-saving, human-machine interaction, assistant driving and human care technologies were illustrated as well.

Abstract:Thermal issues of lithium-ion batteries for automotive application are key factors affecting the performance, safety, life and cost of electric vehicles. In this work, the thermal management systems of three typical electric vehicles were analyzed to identify the importance of the thermal design for the single batteries. Special attention was paid to the review of the thermal modeling, which served as the fundamental method for the thermal design. Finally, the directions for further researches on the thermal modeling and thermal design were summarized.

Abstract:In this paper, a full automatic method was proposed for the segmentation of brain magnetic resonance angiography (MRA) dataset, which improved the technologies of Markov random field (MRF). Existing 3D-MRF models generally faced some challenges including: (1) The parameter initialization of low level MRF model is not accurate; (2) The ordinary neighborhood system cannot deal with local fine vessel structure. Aiming to solve the two problems, the multiscale filtering with threshold analysis and a multi-pattern neighborhood system were proposed, respectively. Such method enabled the MRF model delineating vessels to be as small as two voxels in diameters. In the experiments, the parameters of the low level MRF model were estimated using the expectation maximization algorithm, while the parameters of the high level MRF models were estimated based on the maximum pseudo likelihood algorithm. A set of phantoms and some MRA clinical datasets were used to validate the algorithms, to yield smaller segmentation errors.

Abstract:The medical X-ray image is one of the images most widely applied in clinical applications. Because the lowdose X-ray image needed for imaging is of a low contrast, the X-ray image contrast enhancement is processed before the clinical application. A new algorithm for contrast enhancement of mammographic images was proposed in this paper. The approach was based on the multiscale transform and mathematical morphology. First of all, the Laplacian Gaussian pyramid operator was applied to transform the image into different scale sub-band images. In addition, the high-frequency subimages were equalized by contrast limited adaptive histogram equalization and low-pass sub-images were processed by the mathematical morphology. Finally, the image of enhanced contrast was reconstructed from the Laplacian Gaussian pyramid coefficients of high or low frequencies modified by contrast limited adaptive histogram equalization and mathematical morphology respectively. The enhanced image was processed by a global non-linear operator. The experimental results show that the proposed algorithm is effective for the contrast enhancement of the medical X-ray image. The performances of the proposed algorithm were measured by contrast evaluation criterion for image and contrast improvement index.

Abstract:Falls are the second leading cause of unintentional injury deaths worldwide, so how to prevent falls has become a safety and security problem for elderly people. At present, because the sensing modules of most fall alarms generally only integrate the single 3-axis accelerometer, the measurement accuracy of sensing signals is limited. They can only achieve the alarm of post-fall detection but not the early pre-impact fall recognition in real fall applications. Therefore, an early preimpact fall alarm system based on high-precision inertial sensing units was studied and developed in this paper. A multimodality sensing module embedded with the fall detection algorithm was developed for the early pre-impact fall detection. The module included a 3-axis accelerometer, a 3-axis gyroscope and a 3-axis magnetometer, which could arouse early preimpact fall warnings by a buzzer and a vibrator. Total 81 times of fall experiments from 9 healthy subjects were conducted in simulated fall conditions. The result shows that the detection sensitivity of the system combining the early warning threshold algorithm can reach 98.61% with a specificity of 98.61%, and the average pre-impact lead time is 300 ms. In the future, GPS, GSM electronic modules and wearable protected airbags will be embedded in the system, which will enhance real-time fall protections and timely treatments for the elderly people.

Abstract:Scalespace play an important role in many computer vision tasks. Automatic scale selection is the foundation of multi-scale image analysis, but its performance is still very subjective and empirical. To automatically select the appropriate scale for a particular application, a scale selection model based on information theory was proposed in this paper. The proposed model utilizes the mutual information as a measuring criterion of similarity for the optimal scale selection in multi-scale analysis, with applications to the image denoising and segmentation. Firstly, the multi-scale image smoothing and denoising method based on the morphological operator was studied. This technique does not require the prior knowledge of the noise variance and can effectively eliminate the changes of illumination. Secondly, a clusteringbased unsupervised image segmentation algorithm was developed by recursively pruning the Huffman coding tree. The proposed clustering algorithm can preserve the maximum amount of information at a specific clustering number from the information-theoretical point of view. Finally, for the feasibility of the proposed algorithms, its theoretical properties were analyzed mathematically and its performance was tested through a series of experiments, which demonstrate that it yields the optimal scale for the developed image denoising and segmentation algorithms.

Abstract:The accurate contour delineation of the target and organs at risk (OAR) is essential in treatment planning for image guided radiation therapy. In clinical applications, the contour delineation is often done manually by clinicians, which may be accurate, but time-consuming and tedious for users. Although a lot of automatic contour delineation approaches have been proposed, few of them can fulfill the necessities of applications in terms of accuracy and efficiency. In this work, a novel approach of target delineation was proposed. Target delineation of OARs was achieved by using snake model and multiscale curve editing to obtain promising results. It allows users to quickly improve contours by a simple mouse click. Experimental results demonstrate the effectiveness of the proposed method for clinical target delineations.

Abstract:The uncertain position of lung tumor during radiotherapy compromises the treatment effect. To control the respiratory motion effectively during the radiotherapy of lung cancer without any side effects, a novel control method was introduced in the lung cancer treatment. In order to verify the suggested method, six volunteers were selected with a wide range of distribution of age, weight, and chest circumference. A set of experiments were conducted for each volunteer, under the guidance of a professional hypnotist. All the experiments were repeated in the same environmental condition. The amplitude of respiration was recorded under the normal state and hypnosis, respectively. The mean value and the root mean-square (RMS) of the breathing amplitude were 16.2 mm and 8.6 mm during the hypnosis state, while they were 37.4 mm and 23.9 mm during the normal state. It can be seen that the mean value and the RMS during the hypnosis state were 56.6% and 64.2%, smaller than those during normal state, respectively. Moreover, the stability of the peaks and the similarity of the adjacent wave were also analyzed. The passing ratio of γ index between different cycles during the hypnosis state was 16.4%, higher than that during the normal state. Results demonstrate that the hypnosis intervention can be an alternative way for the respiratory control, which can effectively reduce the respiratory amplitude and increase the stability of the respiratory cycle. The proposed method will find applications in the image guided radiotherapy.

Abstract:The similarity mapping plays an important role in the medical image processing field, including signal investigation, image segmentation and pharmacokinetics of local tissues. Based on tissue similarity mapping (TSM), the concept of dynamic imaging was generalized in this paper, and an improved method was proposed. By the new method, the application fields of TSM can be extended and the image quality can be enhanced dramatically. With 16 series of human brain images acquired from 12-echo T2 * weighted magnetic resonance imaging, these experiments validate that the signal to noise ratio (SNR) from improved method is increased by 3.8 to 17.4 times.