فصلنامه مهندسی پزشکی زیستی
سال چهارم شماره 4 (زمستان 1389)

This paper investigates prototyping an online, low-cost, video based and applicable eye tracker, which is called "Dias Eye Tracker". Disabled people can use the proposed system to communicate with computer. What have made the system different from the other low-cost eye trackers, are the accuracy of gaze estimation, the different application parts of the software and the lightweight wireless hardware, which can be mounted on the user’s head. This paper introduces the software/hardware and the methods of the system. In addition, two methods of pupil tracking have been compared together, and an uncertainty analysis on the mapping function of the system has been done. The performance of the designed eye tracker has been evaluated by analyzing the answers to the three questionnaires, which were filled by disabled people after performing three specific tasks. The results show that the system performs well for interaction with computer.

Many methods for automatic heartbeat classification have been applied and reported in literature, but relatively little number of them concerned with patient independent classification because of the less significant results compared to patient dependent ones. In this work, Reconstructed Phase Space (RPS) theory is used to classify five heartbeat types (Normal, PVC, LBBB, RBBB and PB). In the first and second method, RPS is modeled by the Gaussian mixture model (GMM) and bins, respectively and then classified by classic Bayesian classifier. In the third method, RPS is directly used to train predictor time-delayed neural networks (TDNN) and classified based on minimum prediction error. All three methods highly outperform the results reported before for patient independent heartbeat classification. The best result is achieved using GMM-Bayes method with 92.5% accuracy for patient independent classification.

The aim of this paper is to design a pattern recognition based system to detect P300 component in multi-channel electroencephalogram (EEG) trials. This system has two main blocks, feature extraction and classification. In feature extraction block, in addition to conventional features namely morphological, frequency and wavelet features, some new features included intelligent segmentation, common spatial pattern (CSP) and combined features (CSP + Segmentation) have also been used. Three criteria were used for evaluation and selection of a feature set by choosing a subset of the original features that contains most of essential information. Firstly, a statistical analysis has been applied for evaluating the fitness of each feature in discriminating between target and non target signals. Secondly, each of these six groups of features was evaluated by a Linear Discriminant Analysis (LDA) classifier. Furthermore by using Stepwise Linear Discriminant Analysis (SWLDA), the best set of features was selected. Among these six feature vectors, intelligent segmentation was seen to be most efficient in classification of these signals. In classification phase, two linear classifiers -LDA and SWLDA- were used. The algorithm was described here has tested with dataset II from the BCI competition 2005. In this research, the best result for P300 detection is 97.05% .This result have proven to be more accurate than the results of previous works carried out in this filed.

In this research, we investigated cancerous tissues from several organs of the human body using Raman spectroscopy. Different specimens with different pathologic labels (normal & cancerous) were borrowed from a pathology laboratory, and were investigated using two different Raman spectroscopy systems. Since one of the goals of this investigation was detection of cancer, independent of type of the system, we introduced some algorithms for removing systemic differences from the spectra. Then we removed noise and fluorescence signals using a new wavelet created with LWT. The best classification result was 83% in differentiating between normal and cancerous specimens using the SVM classifier

This paper presents image reconstruction in Diffuse Optical Tomography (DOT) using a high-order finite element method. DOT is a non-invasive imaging modality for visualizing and continuously monitoring tissue and blood oxygenation levels in brain and breast. Image reconstruction in DOT leads to an inverse problem consisting of a forward problem and an iterative algorithm. The inverse problem in DOT systems is ill posed and depends on the accuracy of the forward problem. An accurate model, that describes the light transmission in tissue is required and can increase the spatial resolution. Using first order finite elements in the forward problem, numerical results are converged to the exact solution with increasing the number of elements. However, increasing the number of elements may cause a critical issue in the ill-posed inverse problem. This paper focuses on applying the high-order finite element method without increasing the number of elements, and image reconstruction is accomplished. The forward problem results are compared with analytical solutions. Images of absorbers reconstructed using this method are presented.

One major limitation of walker-supported walking using functional electrical stimulation (FES) in paraplegic subjects is the high energy expenditure and the high upper body effort. Paraplegics should exert high amount of hand force to stabilize the body posture and to compensate lack of the sufficient torques at the lower extremity joints. In this paper, we introduce a 2-D musculoskeletal model of walker-assisted FES-supported walking of paraplegics. Using the developed model and an optimal controller, the stimulation patterns are determined such that the tracking errors of lower joint reference trajectories are minimized and the muscle activations and the handle reaction force (HRF) are reduced. Outputs of the optimal controller are stimulation patterns of the lower body muscles and torque acting on the upper body joints. The results show that the HRF and ground reaction force (GRF) generated by simulation are in agreement with the measured HRF and GRF. Moreover, the results indicate that the simulation-generated stimulation patterns of lower body muscles are in consist with the stimulation patterns reported in the literatures.

Diagnostic follow-up of the brain development during the neonatal period and childhood is an important clinical task. Any disturbance of this process can cause pathological deviations, especially if the baby is born premature. Recent advances in magnetic resonance imaging allow obtaining high-resolution images of the neonatal brain. After segmenting the brains they can be used to reconstruct and model changes occurring during neonatal brain development. In addition such near-realistic model of the head, including the skin, skull and brain can be used to solve the inverse problem of determining the sources of registered signals from electrical brain activity. Although there exist numerous methods and various modeling schemes for adults, these cannot be used directly for neonates due to important differences in morphology. In this review article, neonatal brain atlases are divided into three categories: individual atlases, probabilistic atlases and stochastic atlases. In the following, existing neonatal brain atlases are placed in this classification and their methods of construction are presented. Furthermore, strengths and weaknesses of those neonatal brain atlases are analyzed and finally future research trends in this area are explained.