Dr.Stanislav S.Makhanov | Thammasat University (original) (raw)

Videos by Dr.Stanislav S.Makhanov

We propose a new 5-axis machining technique based on an enhanced vector flow (EVF) The EVF keeps... more We propose a new 5-axis machining technique based on an enhanced vector flow (EVF) The EVF keeps the high-rank vectors unchanged while extending them to unimportant regions, using a diffusion process based on a system of parabolic equations. The resulting enhanced vector field of statistically significant directions (EVFSD) is close to the original VFPD but is characterized by better continuity and regularity.

1 views

A new image processing method in Particle method for segmentation of breast tumors in ultrasou... more A new image processing method in

Particle method for segmentation of breast tumors in ultrasound images, Karunanayake, N., Aimmanee, P., Lohitvisate, W., Makhanov, S.S. Mathematics and Computers in Simulation, 2020, 170, pp. 257-284

3 views

Papers by Dr.Stanislav S.Makhanov

Neural computing & applications, Jun 28, 2024

Medical & biological engineering & computing, Mar 18, 2024

Segmentation of tumors in ultrasound (US) images of the breast is a critical issue in medical ima... more Segmentation of tumors in ultrasound (US) images of the breast is a critical issue in medical imaging. Due to the poor quality of US images and the varying specifications of US machines, segmentation and classification of abnormalities present difficulties even for trained radiologists. The paper aims to introduce a novel AI-based hybrid model for US segmentation that offers high accuracy, requires relatively smaller datasets, and is capable of handling previously unseen data. The software can be used for diagnostics and the US-guided biopsies. A unique and robust hybrid approach that combines deep learning (DL) and multi-agent artificial life (AL) has been introduced. The algorithms are verified on three US datasets. The method outperforms 14 selected state-of-the-art algorithms applied to US images characterized by complex geometry and high level of noise. The paper offers an original classification of the images and tests to analyze the limits of the DL. The model has been trained and verified on 1264 ultrasound images. The images are in the JPEG and PNG formats. The age of the patients ranges from 22 to 73 years. The 14 benchmark algorithms include deformable shapes, edge linking, superpixels, machine learning, and DL methods. The tests use eight-region shape-and contour-based evaluation metrics. The proposed method (DL-AL) produces excellent results in terms of the dice coefficient (region) and the relative Hausdorff distance H 3 (contour-based) as follows: the easiest image complexity level, Dice = 0.96 and H 3 = 0.26; the medium complexity level, Dice = 0.91 and H 3 = 0.82; and the hardest complexity level, Dice = 0.90 and H 3 = 0.84. All other metrics follow the same pattern. The DL-AL outperforms the second best (Unet-based) method by 10-20%. The method has been also tested by a series of unconventional tests. The model was trained on low complexity images and applied to the entire set of images. These results are summarized below. (1) Only the low complexity images have been used for training (68% unknown images): Dice = 0.80 and H 3 = 2.01. (2) The low and the medium complexity images have been used for training (51% unknown images): Dice = 0.86 and H 3 = 1.32. (3) The low, medium, and hard complexity images have been used for training (35% unknown images): Dice = 0.92 and H 3 = 0.76. These tests show a significant advantage of DL-AL over 30%. A video demo illustrating the algorithm is at http:// tinyu rl. com/ mr4ah 687.

Medical & Biological Engineering & Computing, 2024

Segmentation of tumors in ultrasound (US) images of the breast is a critical issue in medical ima... more Segmentation of tumors in ultrasound (US) images of the breast is a critical issue in medical imaging. Due to the poor quality of US images and the varying specifications of US machines, segmentation and classification of abnormalities present difficulties even for trained radiologists. The paper aims to introduce a novel AI-based hybrid model for US segmentation that offers high accuracy, requires relatively smaller datasets, and is capable of handling previously unseen data. The software can be used for diagnostics and the US-guided biopsies. A unique and robust hybrid approach that combines deep learning (DL) and multi-agent artificial life (AL) has been introduced. The algorithms are verified on three US datasets. The method outperforms 14 selected state-of-the-art algorithms applied to US images characterized by complex geometry and high level of noise. The paper offers an original classification of the images and tests to analyze the limits of the DL. The model has been trained and verified on 1264 ultrasound images. The images are in the JPEG and PNG formats. The age of the patients ranges from 22 to 73 years. The 14 benchmark algorithms include deformable shapes, edge linking, superpixels, machine learning, and DL methods. The tests use eight-region shape-and contour-based evaluation metrics. The proposed method (DL-AL) produces excellent results in terms of the dice coefficient (region) and the relative Hausdorff distance H 3 (contour-based) as follows: the easiest image complexity level, Dice = 0.96 and H 3 = 0.26; the medium complexity level, Dice = 0.91 and H 3 = 0.82; and the hardest complexity level, Dice = 0.90 and H 3 = 0.84. All other metrics follow the same pattern. The DL-AL outperforms the second best (Unet-based) method by 10-20%. The method has been also tested by a series of unconventional tests. The model was trained on low complexity images and applied to the entire set of images. These results are summarized below. (1) Only the low complexity images have been used for training (68% unknown images): Dice = 0.80 and H 3 = 2.01. (2) The low and the medium complexity images have been used for training (51% unknown images): Dice = 0.86 and H 3 = 1.32. (3) The low, medium, and hard complexity images have been used for training (35% unknown images): Dice = 0.92 and H 3 = 0.76. These tests show a significant advantage of DL-AL over 30%. A video demo illustrating the algorithm is at http:// tinyu rl. com/ mr4ah 687.

Abstract—Segmentation of tumours in ultrasound (US) images of the breast is a critical problem i... more Abstract—Segmentation of tumours in ultrasound (US) images
of the breast is a critical problem in medical imaging. Due
to the poor quality of US images and varying specifications
of the US machines, the segmentation and classification of the
abnormalities presents difficulties even for trained radiologists.
Nevertheless, the breast US(BUS) remains one of the most
reliable and inexpensive tests for early detection of the breast
cancer. Recent publications show that the automated systems
help radiologists to increase the accuracy of the diagnosis. The
paper offers an algorithm based on Artificial Life (AL) for
breast tumour segmentation. The paper demonstrates that the
AL agents are able to process complex-shaped tumours. The
numerical experiments and tests against preceding models prove
the advantage of AL

International Journal of Production Research, , 2016

We propose a new framework for toolpath generation for five-axis machining of part surfaces repre... more We propose a new framework for toolpath generation for five-axis machining of part surfaces represented by the StereoLithography (STL) format. The framework is based on flattening the STL part and generation of adaptive curvilinear toolpaths. The corresponding cost functions, designed to represent the accuracy and the efficiency of the toolpath, are scalar functions, such as the curvature, kinematic error, rotation angles, machining strip or material removal rate or a vector field when the tool moves along a curvilinear path partly or even entirely aligned with directions considered to be optimal. The adaptive toolpath exploits grid generation methods and biased space-filling curves, combined with adaptation to the boundary and the domain decomposition. The proposed methodology of the adaptive curvilinear toolpath (ACT) has been tested on a variety of STL surfaces, including a case study of STL dental parts. Machining crowns/ implants for four basic types of human teeth, molar, premolar, canine and incisor, has been considered and analysed. The reference methods are the standard iso-parametric path, MasterCam toolpath, and advanced methods of NX9 (former UG). The experiments show that there is no universal sequence of steps applicable to every surface. However, a correct choice of the tools available within the proposed ACT-framework always leads to a substantial improvement of the toolpath, in terms of its length and the machining time.

IOP conference series, Jun 3, 2020

A new algorithm to increase the production rate of a five-axis milling machine through improving ... more A new algorithm to increase the production rate of a five-axis milling machine through improving the coordinates of the to-be-milled points transformed from the workpiece to the machine coordinates is presented and analysed. The method optimizes the cutting path by following a vector field (VF) of optimal directions maximizing the material removal rate (MRR). The algorithm includes grid generation, space filling curves (SFC) and a VF decomposition using rotation invariant complex moments. The case of a radial tool path requires a special treatment called Compact Radial Zigzag (CRZ). To reduce the redundancy, the CRZ is composed of layers with a varying step between the tracks. The combination of the proposed techniques generates tool paths which produce complex shaped Stereolithography (STL) surfaces faster than the conventional methods.

We present a new blur identification procedure based on the hybrid, two-level evolutionary-determ... more We present a new blur identification procedure based on the hybrid, two-level evolutionary-deterministic scheme which allows a dynamic modification of the size and the pattern of the PSF-support. The proposed procedure which comprises a generalization to the case of multiresolutional identification leads to the concept of adaptive interconnection windowing. Numerical simulations demonstrate an overall priority of the proposed identification scheme

Computer Aided Design, Dec 1, 2011

ABSTRACT Kinematics of a particular five-axis milling machine can drastically change the machinin... more ABSTRACT Kinematics of a particular five-axis milling machine can drastically change the machining accuracy. Therefore, the reduction of the kinematics error is an important problem associated with the tool path planning.Our new optimization method employs a closed form of the kinematics error represented as a function of the positions of the cutter contact points. The closed form is derived from the inverse kinematics associated with a particular five-axis machine and obtained through automatic symbolic calculations.The second component of the algorithm is the optimal setup of the part surface on the mounting table employed in an iterative loop with the generation of the cutter contact points.For a prescribed tolerance the proposed optimization allows for substantial reduction in the number of required cutter contact points. The reduction can be significant and may amount to long hours of machining if the machining time at the programmed feed is less than the sampling time of the controller.In turn, when the number of cutter location points is fixed, the error can be substantially reduced. However, this refers to commanded error wherein the dynamics of machine tool are not taken into account.We present an analysis, systematic numerical experiments and results of real cutting (ball nose and flat-end cutters) as an evidence of the efficiency and the accuracy increase produced by the proposed method. We also evaluate the relative contributions of the setup and the point optimization.The method is shown to work with advanced tool path generation techniques proposed earlier such as the adaptive space filling curves.The numerical and machining experiments demonstrate that the proposed procedure outperforms tool paths based on the equi-arclength principle and paths generated by MasterCam 9.

Lecture Notes in Computer Science, 2016

IOP conference series, May 1, 2020

A new algorithm to evaluate the scallops left between consecutive tool tracks after five-axis mac... more A new algorithm to evaluate the scallops left between consecutive tool tracks after five-axis machining of a complex-shaped part surface has been proposed. The algorithm has been developed for the ball-nose cutter. The novelty of the algorithm includes a variable plane to evaluate the effective tool profile and the part surface profile, the orientation of the tool as well as non-linear kinematics of the five-axis machine. The proposed algorithm has been specifically designed for and tested on the industrial Stereo lithography (STL) format representing complex shaped synthetic five-axis parts and a model of a crown of the molar tooth. The procedure has been tested against several modifications of the conventional curvature based method and the sphere intersection method. The ground truth is generated using the solid modeling engine of Vericut 8.2. The algorithm provides a tangible accuracy increase in terms of the average and the maximum error with regard to the reference methods.

Advances in intelligent systems research, 2006

Artificial Life and Robotics, Jan 29, 2022

Retinitis pigmentosa (RP) is a group of genetic disorders, characterized by degeneration of photo... more Retinitis pigmentosa (RP) is a group of genetic disorders, characterized by degeneration of photoreceptor cells which is the main cause of choroidal thinning. It is one of the leading causes of blindness worldwide. Thus, an investigation of choroidal changes is required for a better understanding of disease and diagnosis of RP. In this paper, we propose an automatic technique for measuring the choroidal parameters in optical coherence tomography (OCT) images of eyes with RP. The parameters include the total choroidal area (TCA), luminal area (LA), stromal area (SA), and choroidal thickness (CT). We applied our recently proposed, dense dilated U-Net segmentation model, called ChoroidNET, for segmenting the choroid layer and choroidal vessels for our RP dataset. Choroid segmentation is an important task since the measurement results depend on it. Comparison with other state-of-the-art models shows that ChoroidNET provides a better quantitative and qualitative segmentation of the choroid layer and choroidal vessels. Next, we measure the choroidal parameters based on the segmentation results of ChoroidNET. The proposed method achieves high reliability with an intraclass correlation coefcient (0.961, 0.940, 0.826, 0.916) for TCA, LA, SA, and CT, respectively.

Operations Research Letters, Mar 1, 2018

International Journal of Production Research, Apr 29, 2016

We propose a new framework for toolpath generation for five-axis machining of part surfaces repre... more We propose a new framework for toolpath generation for five-axis machining of part surfaces represented by the StereoLithography (STL) format. The framework is based on flattening the STL part and generation of adaptive curvilinear toolpaths. The corresponding cost functions, designed to represent the accuracy and the efficiency of the toolpath, are scalar functions, such as the curvature, kinematic error, rotation angles, machining strip or material removal rate or a vector field when the tool moves along a curvilinear path partly or even entirely aligned with directions considered to be optimal. The adaptive toolpath exploits grid generation methods and biased space-filling curves, combined with adaptation to the boundary and the domain decomposition. The proposed methodology of the adaptive curvilinear toolpath (ACT) has been tested on a variety of STL surfaces, including a case study of STL dental parts. Machining crowns/implants for four basic types of human teeth, molar, premolar, canine and incisor, has been considered and analysed. The reference methods are the standard iso-parametric path, MasterCam toolpath, and advanced methods of NX9 (former UG). The experiments show that there is no universal sequence of steps applicable to every surface. However, a correct choice of the tools available within the proposed ACT-framework always leads to a substantial improvement of the toolpath, in terms of its length and the machining time.

2018 International Workshop on Advanced Image Technology (IWAIT), 2018

In order to analyze the relationship between ultrasonic signal and tissue structure, accurate ima... more In order to analyze the relationship between ultrasonic signal and tissue structure, accurate image registration is required. However, spatial resolution and image features are different between pathological and ultrasonic images. Thus, this paper proposed an image feature conversion method including downscale process using convolutional neural network. The proposed method was applied to the pathological images and we confirmed that the converted pathological images were similar to the ultrasonic images from visual assessment and image registration was also successfully conducted.

The paper presents a method to increase the production rate of a five axis milling machine throug... more The paper presents a method to increase the production rate of a five axis milling machine through improving how to change the coordinates of the to-be-milled points from the “workpiece” to the “machine” coordinate system, and then deciding the cutting path by maximizing the material removal rate. The method to optimize the tool path is based on the analysis of a vector field of optimal directions which can be pre-computed for a particular machine configuration. The method includes grid generation, space filling curves and vector field decomposition using rotationally invariant moments. The proposed tool path performs the machining 1.2-6 times faster than conventional methods.

Computers in Biology and Medicine, Oct 1, 2020

We propose a new 5-axis machining technique based on an enhanced vector flow (EVF) The EVF keeps... more We propose a new 5-axis machining technique based on an enhanced vector flow (EVF) The EVF keeps the high-rank vectors unchanged while extending them to unimportant regions, using a diffusion process based on a system of parabolic equations. The resulting enhanced vector field of statistically significant directions (EVFSD) is close to the original VFPD but is characterized by better continuity and regularity.

1 views

A new image processing method in Particle method for segmentation of breast tumors in ultrasou... more A new image processing method in

Particle method for segmentation of breast tumors in ultrasound images, Karunanayake, N., Aimmanee, P., Lohitvisate, W., Makhanov, S.S. Mathematics and Computers in Simulation, 2020, 170, pp. 257-284

3 views

Neural computing & applications, Jun 28, 2024

Medical & biological engineering & computing, Mar 18, 2024

Segmentation of tumors in ultrasound (US) images of the breast is a critical issue in medical ima... more Segmentation of tumors in ultrasound (US) images of the breast is a critical issue in medical imaging. Due to the poor quality of US images and the varying specifications of US machines, segmentation and classification of abnormalities present difficulties even for trained radiologists. The paper aims to introduce a novel AI-based hybrid model for US segmentation that offers high accuracy, requires relatively smaller datasets, and is capable of handling previously unseen data. The software can be used for diagnostics and the US-guided biopsies. A unique and robust hybrid approach that combines deep learning (DL) and multi-agent artificial life (AL) has been introduced. The algorithms are verified on three US datasets. The method outperforms 14 selected state-of-the-art algorithms applied to US images characterized by complex geometry and high level of noise. The paper offers an original classification of the images and tests to analyze the limits of the DL. The model has been trained and verified on 1264 ultrasound images. The images are in the JPEG and PNG formats. The age of the patients ranges from 22 to 73 years. The 14 benchmark algorithms include deformable shapes, edge linking, superpixels, machine learning, and DL methods. The tests use eight-region shape-and contour-based evaluation metrics. The proposed method (DL-AL) produces excellent results in terms of the dice coefficient (region) and the relative Hausdorff distance H 3 (contour-based) as follows: the easiest image complexity level, Dice = 0.96 and H 3 = 0.26; the medium complexity level, Dice = 0.91 and H 3 = 0.82; and the hardest complexity level, Dice = 0.90 and H 3 = 0.84. All other metrics follow the same pattern. The DL-AL outperforms the second best (Unet-based) method by 10-20%. The method has been also tested by a series of unconventional tests. The model was trained on low complexity images and applied to the entire set of images. These results are summarized below. (1) Only the low complexity images have been used for training (68% unknown images): Dice = 0.80 and H 3 = 2.01. (2) The low and the medium complexity images have been used for training (51% unknown images): Dice = 0.86 and H 3 = 1.32. (3) The low, medium, and hard complexity images have been used for training (35% unknown images): Dice = 0.92 and H 3 = 0.76. These tests show a significant advantage of DL-AL over 30%. A video demo illustrating the algorithm is at http:// tinyu rl. com/ mr4ah 687.

Medical & Biological Engineering & Computing, 2024

Segmentation of tumors in ultrasound (US) images of the breast is a critical issue in medical ima... more Segmentation of tumors in ultrasound (US) images of the breast is a critical issue in medical imaging. Due to the poor quality of US images and the varying specifications of US machines, segmentation and classification of abnormalities present difficulties even for trained radiologists. The paper aims to introduce a novel AI-based hybrid model for US segmentation that offers high accuracy, requires relatively smaller datasets, and is capable of handling previously unseen data. The software can be used for diagnostics and the US-guided biopsies. A unique and robust hybrid approach that combines deep learning (DL) and multi-agent artificial life (AL) has been introduced. The algorithms are verified on three US datasets. The method outperforms 14 selected state-of-the-art algorithms applied to US images characterized by complex geometry and high level of noise. The paper offers an original classification of the images and tests to analyze the limits of the DL. The model has been trained and verified on 1264 ultrasound images. The images are in the JPEG and PNG formats. The age of the patients ranges from 22 to 73 years. The 14 benchmark algorithms include deformable shapes, edge linking, superpixels, machine learning, and DL methods. The tests use eight-region shape-and contour-based evaluation metrics. The proposed method (DL-AL) produces excellent results in terms of the dice coefficient (region) and the relative Hausdorff distance H 3 (contour-based) as follows: the easiest image complexity level, Dice = 0.96 and H 3 = 0.26; the medium complexity level, Dice = 0.91 and H 3 = 0.82; and the hardest complexity level, Dice = 0.90 and H 3 = 0.84. All other metrics follow the same pattern. The DL-AL outperforms the second best (Unet-based) method by 10-20%. The method has been also tested by a series of unconventional tests. The model was trained on low complexity images and applied to the entire set of images. These results are summarized below. (1) Only the low complexity images have been used for training (68% unknown images): Dice = 0.80 and H 3 = 2.01. (2) The low and the medium complexity images have been used for training (51% unknown images): Dice = 0.86 and H 3 = 1.32. (3) The low, medium, and hard complexity images have been used for training (35% unknown images): Dice = 0.92 and H 3 = 0.76. These tests show a significant advantage of DL-AL over 30%. A video demo illustrating the algorithm is at http:// tinyu rl. com/ mr4ah 687.

Abstract—Segmentation of tumours in ultrasound (US) images of the breast is a critical problem i... more Abstract—Segmentation of tumours in ultrasound (US) images
of the breast is a critical problem in medical imaging. Due
to the poor quality of US images and varying specifications
of the US machines, the segmentation and classification of the
abnormalities presents difficulties even for trained radiologists.
Nevertheless, the breast US(BUS) remains one of the most
reliable and inexpensive tests for early detection of the breast
cancer. Recent publications show that the automated systems
help radiologists to increase the accuracy of the diagnosis. The
paper offers an algorithm based on Artificial Life (AL) for
breast tumour segmentation. The paper demonstrates that the
AL agents are able to process complex-shaped tumours. The
numerical experiments and tests against preceding models prove
the advantage of AL

International Journal of Production Research, , 2016

We propose a new framework for toolpath generation for five-axis machining of part surfaces repre... more We propose a new framework for toolpath generation for five-axis machining of part surfaces represented by the StereoLithography (STL) format. The framework is based on flattening the STL part and generation of adaptive curvilinear toolpaths. The corresponding cost functions, designed to represent the accuracy and the efficiency of the toolpath, are scalar functions, such as the curvature, kinematic error, rotation angles, machining strip or material removal rate or a vector field when the tool moves along a curvilinear path partly or even entirely aligned with directions considered to be optimal. The adaptive toolpath exploits grid generation methods and biased space-filling curves, combined with adaptation to the boundary and the domain decomposition. The proposed methodology of the adaptive curvilinear toolpath (ACT) has been tested on a variety of STL surfaces, including a case study of STL dental parts. Machining crowns/ implants for four basic types of human teeth, molar, premolar, canine and incisor, has been considered and analysed. The reference methods are the standard iso-parametric path, MasterCam toolpath, and advanced methods of NX9 (former UG). The experiments show that there is no universal sequence of steps applicable to every surface. However, a correct choice of the tools available within the proposed ACT-framework always leads to a substantial improvement of the toolpath, in terms of its length and the machining time.

IOP conference series, Jun 3, 2020

A new algorithm to increase the production rate of a five-axis milling machine through improving ... more A new algorithm to increase the production rate of a five-axis milling machine through improving the coordinates of the to-be-milled points transformed from the workpiece to the machine coordinates is presented and analysed. The method optimizes the cutting path by following a vector field (VF) of optimal directions maximizing the material removal rate (MRR). The algorithm includes grid generation, space filling curves (SFC) and a VF decomposition using rotation invariant complex moments. The case of a radial tool path requires a special treatment called Compact Radial Zigzag (CRZ). To reduce the redundancy, the CRZ is composed of layers with a varying step between the tracks. The combination of the proposed techniques generates tool paths which produce complex shaped Stereolithography (STL) surfaces faster than the conventional methods.

We present a new blur identification procedure based on the hybrid, two-level evolutionary-determ... more We present a new blur identification procedure based on the hybrid, two-level evolutionary-deterministic scheme which allows a dynamic modification of the size and the pattern of the PSF-support. The proposed procedure which comprises a generalization to the case of multiresolutional identification leads to the concept of adaptive interconnection windowing. Numerical simulations demonstrate an overall priority of the proposed identification scheme

Computer Aided Design, Dec 1, 2011

ABSTRACT Kinematics of a particular five-axis milling machine can drastically change the machinin... more ABSTRACT Kinematics of a particular five-axis milling machine can drastically change the machining accuracy. Therefore, the reduction of the kinematics error is an important problem associated with the tool path planning.Our new optimization method employs a closed form of the kinematics error represented as a function of the positions of the cutter contact points. The closed form is derived from the inverse kinematics associated with a particular five-axis machine and obtained through automatic symbolic calculations.The second component of the algorithm is the optimal setup of the part surface on the mounting table employed in an iterative loop with the generation of the cutter contact points.For a prescribed tolerance the proposed optimization allows for substantial reduction in the number of required cutter contact points. The reduction can be significant and may amount to long hours of machining if the machining time at the programmed feed is less than the sampling time of the controller.In turn, when the number of cutter location points is fixed, the error can be substantially reduced. However, this refers to commanded error wherein the dynamics of machine tool are not taken into account.We present an analysis, systematic numerical experiments and results of real cutting (ball nose and flat-end cutters) as an evidence of the efficiency and the accuracy increase produced by the proposed method. We also evaluate the relative contributions of the setup and the point optimization.The method is shown to work with advanced tool path generation techniques proposed earlier such as the adaptive space filling curves.The numerical and machining experiments demonstrate that the proposed procedure outperforms tool paths based on the equi-arclength principle and paths generated by MasterCam 9.

Lecture Notes in Computer Science, 2016

IOP conference series, May 1, 2020

A new algorithm to evaluate the scallops left between consecutive tool tracks after five-axis mac... more A new algorithm to evaluate the scallops left between consecutive tool tracks after five-axis machining of a complex-shaped part surface has been proposed. The algorithm has been developed for the ball-nose cutter. The novelty of the algorithm includes a variable plane to evaluate the effective tool profile and the part surface profile, the orientation of the tool as well as non-linear kinematics of the five-axis machine. The proposed algorithm has been specifically designed for and tested on the industrial Stereo lithography (STL) format representing complex shaped synthetic five-axis parts and a model of a crown of the molar tooth. The procedure has been tested against several modifications of the conventional curvature based method and the sphere intersection method. The ground truth is generated using the solid modeling engine of Vericut 8.2. The algorithm provides a tangible accuracy increase in terms of the average and the maximum error with regard to the reference methods.

Advances in intelligent systems research, 2006

Artificial Life and Robotics, Jan 29, 2022

Retinitis pigmentosa (RP) is a group of genetic disorders, characterized by degeneration of photo... more Retinitis pigmentosa (RP) is a group of genetic disorders, characterized by degeneration of photoreceptor cells which is the main cause of choroidal thinning. It is one of the leading causes of blindness worldwide. Thus, an investigation of choroidal changes is required for a better understanding of disease and diagnosis of RP. In this paper, we propose an automatic technique for measuring the choroidal parameters in optical coherence tomography (OCT) images of eyes with RP. The parameters include the total choroidal area (TCA), luminal area (LA), stromal area (SA), and choroidal thickness (CT). We applied our recently proposed, dense dilated U-Net segmentation model, called ChoroidNET, for segmenting the choroid layer and choroidal vessels for our RP dataset. Choroid segmentation is an important task since the measurement results depend on it. Comparison with other state-of-the-art models shows that ChoroidNET provides a better quantitative and qualitative segmentation of the choroid layer and choroidal vessels. Next, we measure the choroidal parameters based on the segmentation results of ChoroidNET. The proposed method achieves high reliability with an intraclass correlation coefcient (0.961, 0.940, 0.826, 0.916) for TCA, LA, SA, and CT, respectively.

Operations Research Letters, Mar 1, 2018

International Journal of Production Research, Apr 29, 2016

We propose a new framework for toolpath generation for five-axis machining of part surfaces repre... more We propose a new framework for toolpath generation for five-axis machining of part surfaces represented by the StereoLithography (STL) format. The framework is based on flattening the STL part and generation of adaptive curvilinear toolpaths. The corresponding cost functions, designed to represent the accuracy and the efficiency of the toolpath, are scalar functions, such as the curvature, kinematic error, rotation angles, machining strip or material removal rate or a vector field when the tool moves along a curvilinear path partly or even entirely aligned with directions considered to be optimal. The adaptive toolpath exploits grid generation methods and biased space-filling curves, combined with adaptation to the boundary and the domain decomposition. The proposed methodology of the adaptive curvilinear toolpath (ACT) has been tested on a variety of STL surfaces, including a case study of STL dental parts. Machining crowns/implants for four basic types of human teeth, molar, premolar, canine and incisor, has been considered and analysed. The reference methods are the standard iso-parametric path, MasterCam toolpath, and advanced methods of NX9 (former UG). The experiments show that there is no universal sequence of steps applicable to every surface. However, a correct choice of the tools available within the proposed ACT-framework always leads to a substantial improvement of the toolpath, in terms of its length and the machining time.

2018 International Workshop on Advanced Image Technology (IWAIT), 2018

In order to analyze the relationship between ultrasonic signal and tissue structure, accurate ima... more In order to analyze the relationship between ultrasonic signal and tissue structure, accurate image registration is required. However, spatial resolution and image features are different between pathological and ultrasonic images. Thus, this paper proposed an image feature conversion method including downscale process using convolutional neural network. The proposed method was applied to the pathological images and we confirmed that the converted pathological images were similar to the ultrasonic images from visual assessment and image registration was also successfully conducted.

The paper presents a method to increase the production rate of a five axis milling machine throug... more The paper presents a method to increase the production rate of a five axis milling machine through improving how to change the coordinates of the to-be-milled points from the “workpiece” to the “machine” coordinate system, and then deciding the cutting path by maximizing the material removal rate. The method to optimize the tool path is based on the analysis of a vector field of optimal directions which can be pre-computed for a particular machine configuration. The method includes grid generation, space filling curves and vector field decomposition using rotationally invariant moments. The proposed tool path performs the machining 1.2-6 times faster than conventional methods.

Computers in Biology and Medicine, Oct 1, 2020

Communications in Statistics, Jan 10, 2021

We introduce the global recovery time of Gaver’s parallel system supervised by a safety unit. The... more We introduce the global recovery time of Gaver’s parallel system supervised by a safety unit. The entire system is attended by two heterogeneous repairmen. Our methodology is based on the theory of...