Ali Sadeghi Naini

Statistical finite element method for real-time tissue mechanics analysis

The finite element (FE) method can accurately calculate tissue deformation. However, its low speed renders it ineffective for many biomedical applications involving real-time data processing. To accelerate FE analysis, we introduce a novel tissue mechanics simulation technique. This technique is suitable for real-time estimation of tissue deformation of specific organs, which is required in computer-aided diagnostic or therapeutic procedures. In this method, principal component analysis is used to describe each organ shape and its corresponding FE field for a pool of patients by a small number of weight factors. A mapping function is developed to relate the parameters of organ shape to their FE field counterpart. We show that irrespective of the complexity of the tissues constitutive law or its loading conditions, the proposed technique is highly accurate and fast in estimating the FE field. Average deformation errors of less than 2% demonstrate the accuracy of the proposed technique.

CT-Enhanced Ultrasound Image of a Totally Deflated Lung for Image-Guided Minimally Invasive Tumor Ablative Procedures

A technique is proposed to enhance the quality of intraoperative ultrasound (US) images of a deflated lung undergoing minimally invasive tumor ablative procedure. Since US images are very sensitive to residual air remaining in deflated lung, lung US images have very poor quality, and hence, are not appropriate for image-guided procedures. Therefore, a reliable and high-quality intraoperative image of the lung is a paramount necessity for tumor localization and fusion with real-time navigation data during such procedures. The proposed technique employs information of a deflated lungs computed tomography (CT) image constructed preoperatively in order to enhance those of the intraoperative US images. The enhancement is performed via two concurrent registration processes. The output is an enhanced US image of the deflated lung oriented and positioned accurately within its preoperative CT counterpart. Ex vivo experiments were conducted to evaluate the performance of the proposed technique. The obtained results indicate that very considerable improvement was achieved in the quality of the input intraoperative US images of the deflated lung.

CT image construction of a totally deflated lung using deformable model extrapolation

PURPOSE A novel technique is proposed to construct CT image of a totally deflated lung from a free-breathing 4D-CT image sequence acquired preoperatively. Such a constructed CT image is very useful in performing tumor ablative procedures such as lung brachytherapy. Tumor ablative procedures are frequently performed while the lung is totally deflated. Deflating the lung during such procedures renders preoperative images ineffective for targeting the tumor. Furthermore, the problem cannot be solved using intraoperative ultrasound (U.S.) images because U.S. images are very sensitive to small residual amount of air remaining in the deflated lung. One possible solution to address these issues is to register high quality preoperative CT images of the deflated lung with their corresponding low quality intraoperative U.S. images. However, given that such preoperative images correspond to an inflated lung, such CT images need to be processed to construct CT images pertaining to the lungs deflated state. METHODS To obtain the CT images of deflated lung, we present a novel image construction technique using extrapolated deformable registration to predict the deformation the lung undergoes during full deflation. The proposed construction technique involves estimating the lungs air volume in each preoperative image automatically in order to track the respiration phase of each 4D-CT image throughout a respiratory cycle; i.e., the technique does not need any external marker to form a respiratory signal in the process of curve fitting and extrapolation. The extrapolated deformation field is then applied on a preoperative reference image in order to construct the totally deflated lungs CT image. The technique was evaluated experimentally using ex vivo porcine lung. RESULTS The ex vivo lung experiments led to very encouraging results. In comparison with the CT image of the deflated lung we acquired for the purpose of validation, the constructed CT image was very similar. The intensity mean absolute difference between these two images was calculated to be at 1%. Tumor center as well as a number of anatomical fiducial markers were traced in different corresponding slices of the two images. The average misalignment obtained for the constructed CT image was (0.64, 0.39, 0.11) mm, which indicates a very desirable accuracy for lung brachytherapy applications. CONCLUSIONS The image construction accuracy obtained in this research is suitable for intraoperative tasks; e.g., tumor localization and fusing with real time navigation data in lung brachytherapy. These applications involve image registration with intraoperative U.S. images in order to enhance their poor quality. The proposed technique is also useful for preoperative tasks such as planning of lung brachytherapy treatment.

Estimation of Lung's Air Volume and Its Variations Throughout Respiratory CT Image Sequences

A respiratory image-sequence-segmentation technique is introduced based on a novel image-sequence analysis. The proposed technique is capable of segmenting the lungs air and its soft tissues followed by estimating the lungs air volume and its variations throughout the image sequence. Accurate estimation of these two parameters is very important in many applications related to lung disease diagnosis and treatment systems (e.g., brachytherapy), where the parameters are either the variables of interest themselves or are dependent/independent variables. The concept of the proposed technique involves using the image sequences combined histogram to obtain a reasonable initial guess for the lungs air segmentation thresholds. This is followed by an optimization process to find the optimum threshold values that best satisfy the lungs air mass conservation and tissue incompressibility principles. These threshold values are consequently applied to estimate the lungs air volume and its variations throughout respiratory Computed Tomography (CT) image sequences. Ex vivo experiments were conducted on porcine left lungs in order to demonstrate the performance of the proposed technique. The proposed method was initially validated using a breath-hold CT image sequence with known air volumes inside the lung, where results show that the proposed technique outperforms single-histogram-based methods. This was followed by demonstrating the proposed techniques application in a 4-D-CT respiratory sequence, where the air volume inside the lung was unknown. Consistency of the obtained results in the latter experiment with tissue near incompressibility principle was validated. The results indicate a very good ability of the proposed method for estimating the lungs air volume and its variations in a respiratory image sequence.

MODELLING THE ADAPTATION OF BUSINESS CONTINUITY PLANNING BY BUSINESSES USING NEURAL NETWORKS

SUMMARY Business continuity planning is an important element of business continuity management and is regarded as a fundamental step towards reducing the negative impacts of business disruptions caused by internal and external hazardous events. Many businesses are not prepared for such events, and very few studies have tried to examine and model the factors that contribute to business continuity management planning by various companies. In this paper we propose and develop a feed-forward neural network for modelling businesses continuity planning by businesses based on a dataset of 283 businesses operating in the Greater Toronto Area in Ontario, Canada. The fully connected neural network applied was trained on 65 % of the dataset records using different subsets of input variables. In order to preserve the generalization ability of the trained network, 15 % of the dataset records were used as a validation set for early stopping during the networks training process. Prediction capability of the trained networks was evaluated on 20 % and never-seen records of the dataset. The classification ability of the networks was then analysed using receiver operating characteristic and detection error trade-off curves, where the results obtained were promising. The equal error rate for the best models was 12 %, which reflects a very good accuracy of these models in predicting the existence of business continuity planning for a generic company. Copyright

A. ASGARY AND A. SADEGHI NAINI

SUMMARY Business continuity planning is an important element of business continuity management and is regarded as a fundamental step towards reducing the negative impacts of business disruptions caused by internal and external hazardous events. Many businesses are not prepared for such events, and very few studies have tried to examine and model the factors that contribute to business continuity management planning by various companies. In this paper we propose and develop a feed-forward neural network for modelling businesses continuity planning by businesses based on a dataset of 283 businesses operating in the Greater Toronto Area in Ontario, Canada. The fully connected neural network applied was trained on 65 % of the dataset records using different subsets of input variables. In order to preserve the generalization ability of the trained network, 15 % of the dataset records were used as a validation set for early stopping during the networks training process. Prediction capability of the trained networks was evaluated on 20 % and never-seen records of the dataset. The classification ability of the networks was then analysed using receiver operating characteristic and detection error trade-off curves, where the results obtained were promising. The equal error rate for the best models was 12 %, which reflects a very good accuracy of these models in predicting the existence of business continuity planning for a generic company. Copyright

CT image construction of the lung in a totally deflated mode

A novel technique is proposed to construct CT images of the lung in a totally deflated mode using non-rigid registration and extrapolation. This CT image would be very useful in performing tumor ablative procedures (such as brachytherapy) for the treatment of lung cancer. This is because during such procedures the target lung is almost completely deflated whereas pre-operative images are acquired while the lung is partially inflated. This makes pre-operative images very inaccurate. Given that Ultrasound (US) imaging is very sensitive to residual air in a deflated lung, it is not an effective intra-operative imaging modality by itself. One possible approach for image guided lung brachytherapy is registering low quality intra-operative ultrasound images to high quality lung CT image of the deflated lung constructed using the proposed technique. The technique was applied to an ex-vivo porcine lung and the preliminary results were found to be very encouraging.

Lung tumor motion prediction during lung brachytherapy using finite element model

A biomechanical model is proposed to predict deflated lung tumor motion caused by diaphragm respiratory motion. This model can be very useful for targeting the tumor in tumor ablative procedures such as lung brachytherapy. To minimize motion within the target lung, these procedures are performed while the lung is deflated. However, significant amount of tissue deformation still occurs during respiration due to the diaphragm contact forces. In the absence of effective realtime image guidance, biomechanical models can be used to estimate tumor motion as a function of diaphragms position. To develop this model, Finite Element Method (FEM) was employed. To demonstrate the concept, we conducted an animal study of an ex-vivo porcine deflated lung with a tumor phantom. The lung was deformed by compressing a diaphragm mimicking cylinder against it. Before compression, 3D-CT image of this lung was acquired, which was segmented and turned into FE mesh. The lung tissue was modeled as hyperelastic material with a contact loading to calculate the lung deformation and tumor motion during respiration. To validate the results from FE model, the motion of a small area on the surface close to the tumor was tracked while the lung was being loaded by the cylinder. Good agreement was demonstrated between the experiment results and simulation results. Furthermore, the impact of tissue hyperelastic parameters uncertainties in the FE model was investigated. For this purpose, we performed in-silico simulations with different hyperelastic parameters. This study demonstrated that the FEM was accurate and robust for tumor motion prediction.

A totally deflated lung's CT image construction by means of extrapolated deformable registration

A novel technique is proposed to construct CT image of a totally deflated lung using breath-hold lungs preoperative CT images acquired during respiration. Such a constructed CT image is very useful in tumor targeting during tumor ablative procedures such as lung brachytherapy used for lung cancer treatment. To minimize motion within the target lung, tumor ablative procedures are frequently performed while the lung is totally deflated. Deflating the lung during such procedures renders pre-operative images ineffective for tumor targeting, because those images correspond to the lung while it is partially inflated. Furthermore, the problem cannot be solved using intra-operative Ultrasound (US) images. This is because the quality of lung US images degrades substantially as a result of the residual air inside the deflated lung, thus it is not an effective intra-operative imaging modality by itself. One possible approach for image-guided lung brachytherapy is to register high quality preoperative CT images of the deflated lung with their corresponding low quality intra-operative US images. To obtain the CT images of deflated lung, a novel image construction technique is presented. The proposed technique was implemented using two deformable registration methods: multi-resolution B-spline and multi-resolution demons. The technique was applied to ex vivo porcine lungs where results obtained were found to be very encouraging.

MODELLING THE ADAPTATION OF BUSINESS CONTINUITY PLANNING BY BUSINESSES USING NEURAL NETWORKS: MODELLING BUSINESS CONTINUITY PLANNING USING NEURAL NETWORKS

SUMMARY Business continuity planning is an important element of business continuity management and is regarded as a fundamental step towards reducing the negative impacts of business disruptions caused by internal and external hazardous events. Many businesses are not prepared for such events, and very few studies have tried to examine and model the factors that contribute to business continuity management planning by various companies. In this paper we propose and develop a feed-forward neural network for modelling businesses continuity planning by businesses based on a dataset of 283 businesses operating in the Greater Toronto Area in Ontario, Canada. The fully connected neural network applied was trained on 65 % of the dataset records using different subsets of input variables. In order to preserve the generalization ability of the trained network, 15 % of the dataset records were used as a validation set for early stopping during the networks training process. Prediction capability of the trained networks was evaluated on 20 % and never-seen records of the dataset. The classification ability of the networks was then analysed using receiver operating characteristic and detection error trade-off curves, where the results obtained were promising. The equal error rate for the best models was 12 %, which reflects a very good accuracy of these models in predicting the existence of business continuity planning for a generic company. Copyright