Papers by Amir hossein Foruzan
電子情報通信学会技術研究報告. PRMU, パターン認識・メディア理解, May 12, 2011
電子情報通信学会技術研究報告. PRMU, パターン認識・メディア理解, May 6, 2010
ABSTRACT Recently a growing interest has been seen in minimally invasive treatments with open con... more ABSTRACT Recently a growing interest has been seen in minimally invasive treatments with open configuration magnetic resonance (Open-MR) scanners. Because of the lower magnetic field (0.5T) and various different surgical conditions, sometimes tumors can not be visualized clearly on Open-MR volumes. Combining of CT volumes acquired before surgery, it is possible to identify the tumor's location by application of registration techniques. In this paper, we proposed a non-rigid registration method combined with a semi-automatic liver segmentation method for MR-Guided Liver Cancer Surgery. We first propose a robust method using K-means clustering and graph cut to segment liver from low contrast open MR images and then a free-form deformation based non-rigid registration method is applied to the segmented livers.
電子情報通信学会技術研究報告. MI, 医用画像, Jan 18, 2008
電子情報通信学会ソサイエティ大会講演論文集, Sep 9, 2014
International Conference on Software Engineering, Jun 23, 2010
Since the medical training samples are very limited, it is difficult to construct a statistical s... more Since the medical training samples are very limited, it is difficult to construct a statistical shape model with good generalization using few samples. In this paper, we propose a novel statistical shape modeling method using 2D PCA. The 3D shape is represented as a matrix by spherical parameterization. The experiments showed that our proposed method can reconstruct statistical shape model
IEICE Technical Report; IEICE Tech. Rep., May 21, 2009
In this paper, we present an algorithm to segment the liver in low-contrast CT images. As the fir... more In this paper, we present an algorithm to segment the liver in low-contrast CT images. As the first step of our algorithm, we define a search range for the liver boundary. Then, the EM algorithm is utilized to estimate parameters of a 'Gaussian Mixture' model that conforms to the intensity distribution of the liver. Using the statistical parameters of the intensity distribution, we introduce a new thresholding technique to classify image pixels. We assign a distance feature vectors to each pixel and segment the liver by a K-means clustering scheme. This initial boundary of the liver is conditioned by the Fourier transform. Then, a Geodesic Active Contour algorithm uses the boundaries to find the final surface. The novelty in our method is the proper selection and combination of sub-algorithms so as to find the border of an object in a low-contrast image. The number of parameters in the proposed method is low and the parameters have a low range of variations. We applied our method to 30 datasets including normal and abnormal cases of low-contrast/high-contrast images and it was extensively evaluated both quantitatively and qualitatively. Minimum of Dice similarity measures of the results is 0.89. Assessment of the results proves the potential of the proposed method for segmentation in low-contrast images.
In computational anatomy, statistical shape model (SSM) is used for the quantitative evaluation o... more In computational anatomy, statistical shape model (SSM) is used for the quantitative evaluation of variations in the shapes of different organs. This paper focuses on the construction of a SSM of the liver and its application to computer-assisted diagnosis of cirrhosis. We prove the potential application of SSMs in the classification of normal and cirrhotic livers. In constructing a SSM
Biomedical Engineering: Applications, Basis and Communications, Jun 13, 2022
Successful treatment of a patient depends on the accurate determination of the disease type. The ... more Successful treatment of a patient depends on the accurate determination of the disease type. The advent of big data facilitates the retrieving of medical images and helps physicians in reliable diagnoses using content-based medical image retrieval systems (CBMIR). They consist of a feature extraction module and a distance metric. The extracted textural or deep-based features identify different types of diseases. In the proposed retrieval algorithm, we use the gray level cooccurrence matrix as the common textural characteristics and integrate them with semantic attributes. The semantic features are the geometric characteristics of the tumor that a radiologist employ to distinguish between benign and malignant tumors. These high-level attributes include the Euler number, margin smoothness, and the aspect ratio of the lesion’s size. We used the Minkowski distance measure for computing the similarity of images and applied the proposed algorithm to 200 CT-scan data containing lung lesions obtained from the LIDC database. The types of lesions were benign and malignant. Employing an ablation study, we proved the effectiveness of the semantic feature. The precision of the retrieval results is 93% which is promising compared to recent studies. In the future, we plan to define other kinds of semantic attributes to distinguish stages 1–5 of lung tumors as well.
International Journal of Image and Graphics, Jul 14, 2021
Accurate delineation of the prostate in MR images is an essential step for treatment planning and... more Accurate delineation of the prostate in MR images is an essential step for treatment planning and volume estimation of the organ. Prostate segmentation is a challenging task due to its variable size and shape. Moreover, neighboring tissues have a low-contrast with the prostate. We propose a robust and precise automatic algorithm to define the prostate’s boundaries in MR images in this paper. First, we find the prostate’s ROI by a deep neural network and decrease the input image’s size. Next, a dynamic multi-atlas-based approach obtains the initial segmentation of the prostate. A watershed algorithm improves the initial segmentation at the next stage. Finally, an SSM algorithm keeps the result in the domain of allowable prostate shapes. The quantitative evaluation of 74 prostate volumes demonstrated that the proposed method yields a mean Dice coefficient of [Formula: see text]. In comparison with recent researches, our algorithm is robust against shape and size variations.
Journal of Flow Visualization and Image Processing, 2021
In this paper, we proposed a method for liver segmentation in CT-scan images that is based on geo... more In this paper, we proposed a method for liver segmentation in CT-scan images that is based on geometric active contours. After reading an input image, we obtain an initial segmentation by an intensity-based technique. Then, we apply a dilation filter on the initial surface to obtain an incremental narrow volume around it. We divide this volume into smaller parts and we prepare the corresponding velocity field in each part individually. The initial surface evolves under the velocity field and this process iterates until convergence. We applied our method on 30 CT-scan datasets including healthy/patient people, and contrast-enhanced/low-contrast images. The average Dice and Jaccard indices were 0.91 and 0.84 respectively. Compared to the STACS algorithm, we improved Dice and Jaccard indices by 0.63 and 0.47.
In this paper, we incorporate a locally estimated appearance model to enrich the data term of the... more In this paper, we incorporate a locally estimated appearance model to enrich the data term of the graph-cuts algorithm. It balances between data term and smoothing terms in order to extract small liver vessels. We estimate stochastic parameters of vessel and liver tissues using the MIP image. Medial axes of the vessels are then enhanced by a multi-scale filter. The skeleton of the axes are used to prepare the appearance model which is employed to prepare s-link weights in the graph-cuts algorithm. We evaluated the proposed method quantitatively using public synthetic data and qualitatively using clinical images. We obtained an average Dice measure of 0.93 which was comparable to recent researches. We achieved segmentation of liver vessels up to the fourth order too.
Advances in computer science research, 2015
Probabilistic atlases based on human anatomy structure have been widely used for organ segmentati... more Probabilistic atlases based on human anatomy structure have been widely used for organ segmentation, which is used as a prior probability in a Bayes framework. The challenge is how to register the probabilistic atlas to the patient volume. Taking these into consideration, we propose a template matching framework based on the probabilistic atlas for spleen segmentation. Firstly, we find a bounding box of the spleen based on human anatomical localization, which is the statistical geometric location of spleens. Then, the probabilistic atlas is used as a template to find the spleen in this bounding box by using template matching technology. We apply our method into 60 datasets including normal and pathological cases. The Dice/Tanimoto volume overlaps are 0.922/0.857, the root-meansquared error (RMSE) is 1.992 mm. The algorithm is robust to segment normal and abnormal spleens, such as the presence of tumors and large morphological changes. Meanwhile, our proposed method was compared with conventional atlas-based methods. Results demonstrate that segmentation accuracy improved using our method.
International Conference on Software Engineering, Jun 23, 2010
ABSTRACT In computational anatomy, statistical shape model is used for quantitative evaluation of... more ABSTRACT In computational anatomy, statistical shape model is used for quantitative evaluation of the variations of an organ shape. This paper is focused on construction of Statistical Shape Model of the liver and its application to computer assisted diagnosis. We prove the potential application of statistical shape models in classification of normal and cirrhosis livers. First, statistical shape model of liver is constructed. Then the coefficients of the model are used to recognize whether liver is normal or abnormal.
Iranian Journal of Science and Technology Transaction A-science, Apr 8, 2019
Previous studies show that ultrasound has a critical role in the synthesis of nanoparticles. In t... more Previous studies show that ultrasound has a critical role in the synthesis of nanoparticles. In this paper, we compare the synthesis of MnO 2 nanoparticles in the presence and absence of ultrasonic waves. The reaction efficiency is measured by the back titration method. The purity, phase, and morphology of the prepared samples are analyzed by X-ray diffraction (XRD) and field-emission scanning electron microscopy (FESEM). The surface area and total pore volume are measured by Brunauer-Emmett-Teller (BET) analysis. The results showed that the reaction efficiency reached 95% in the presence of ultrasonic waves during only 20 min, while it needed more than 6 h to reach such efficiency in the absence of ultrasonic waves. The XRD analysis and FESEM images showed that both samples were MnO 2 nanoparticles and their phase was α. The increased synthesis rate is due to the production of free radicals during water sonolysis. Nitrogen adsorption-desorption isotherms displayed mesoporous nanoparticles with a pore size in the range of 2-10 nm for both samples. The surface area was 28.092 m 2 /g for samples under ultrasonic irradiation and 77.533 m 2 /g for other samples. A decrease of 2.75 times in the surface area was demonstrated for samples under ultrasonic irradiation. Also, the total pore volume of samples under ultrasonic irradiation was 1.85 times less than that of other samples. The reduction in the surface area and total pore volume is due to the role of the shock wave phenomenon in ultrasound irradiation. The shock wave increases the collision among nanoparticles and the adhesion among them. Based on the analysis of FESEM images, a new index is introduced describing the cohesiveness of the samples. This measure is compatible with the results of the BET analysis and shows increased adhesion among nanoparticles.
Measurement, Mar 1, 2020
The ultrasonic irradiation of a liquid generates acoustic bubbles. Collapsed bubbles in the acous... more The ultrasonic irradiation of a liquid generates acoustic bubbles. Collapsed bubbles in the acoustic pressure create a hot spot condition consisting of high temperature and pressure. Determining the size of the bubbles is vital in the characterization of the corresponding ultrasound wave and the hot spot condition. In this paper, we estimate the distribution of the radius of acoustic bubbles by an image processing technique based on the Principal Component Analysis. We automatically measure the velocity of the bubbles, calculate their radius, and compare the percentage of small/large bubbles and bubble clusters from their radius distribution. We performed several experiments using ultrasonic horn tips of various diameters and ultrasound powers. The results showed that the mean bubble size for a 3 mm tip is 75.66 mm and 82.36 mm in 5 W and 23 W radiation powers, while for a 20 mm tip, the size is 98.3 mm and 109.06 mm for 80 W and 260 W powers, respectively.
Representation of soft tissues in virtual reality environments has been focused by researchers wi... more Representation of soft tissues in virtual reality environments has been focused by researchers with applications including training medical students and surgeons, treatment planning, monitoring and telesurgery. A major challenge of current modeling schemes such as Boundary Element, Finite Element, and Mass-Spring Models is to deal with volume preserving. Another challenge is the complexity of a model which results in a more realistic visualization; however, it increases computational cost. In this paper, we propose a Mass-Spring model to represent liver volume. It contains a series of multi-scale surface meshes with interconnections between the models and therefore it is considered as a volumetric mesh model. To preserve the volume of the gland, an external force is transmitted from the surface to internal meshes. By designing a specific data structure to hold coordinates of mesh points, we are able to render mesh movement in nearly real-time using conventional CPU architectures. Localization of the external force is adjusted by the penetration depth parameter. Qualitative evaluation of the results revealed the promising performance of the proposed model. The stability of our Mass-Spring model under large deformation is another novelty of our method too.
Statistical shape model (SSM) is to model the shape variation of an object. In this paper, we pro... more Statistical shape model (SSM) is to model the shape variation of an object. In this paper, we propose an efficient shape representation method and a new 2D-PCA based statistical shape modeling. In our proposed method, we used the radii of these surface points as shape feature instead of their coordinates, and the shape is represented by a 2D matrices. We
Uploads
Papers by Amir hossein Foruzan