The Experts below are selected from a list of 51 Experts worldwide ranked by ideXlab platform
Nuo Tong - One of the best experts on this subject based on the ideXlab platform.
-
fully automatic multi organ Segmentation for head and neck cancer radiotherapy using shape representation model constrained fully convolutional neural Networks
Medical Physics, 2018Co-Authors: Nuo Tong, Shuyuan Yang, Dan Ruan, K ShengAbstract:PURPOSE: Intensity modulated radiation therapy (IMRT) is commonly employed for treating head and neck (HN (b) the pre-trained SRM with fixed parameters were used to constrain the FCNN training. The combined Segmentation Network was then used to delineate nine OARs including the brainstem, optic chiasm, mandible, optical nerves, parotids, and submandibular glands on unseen H&N CT images. Twenty-two and 10 H&N CT scans provided by the Public Domain Database for Computational Anatomy (PDDCA) were utilized for training and validation, respectively. Dice similarity coefficient (DSC), positive predictive value (PPV), sensitivity (SEN), average surface distance (ASD), and 95% maximum surface distance (95%SD) were calculated to quantitatively evaluate the Segmentation accuracy of the proposed method. The proposed method was compared with an active appearance model that won the 2015 MICCAI H&N Segmentation Grand Challenge based on the same dataset, an atlas method and a deep learning method based on different patient datasets. RESULTS: An average DSC = 0.870 (brainstem), DSC = 0.583 (optic chiasm), DSC = 0.937 (mandible), DSC = 0.653 (left optic nerve), DSC = 0.689 (right optic nerve), DSC = 0.835 (left parotid), DSC = 0.832 (right parotid), DSC = 0.755 (left submandibular), and DSC = 0.813 (right submandibular) were achieved. The Segmentation results are consistently superior to the results of atlas and statistical shape based methods as well as a patch-wise convolutional neural Network method. Once the Networks are trained off-line, the average time to segment all 9 OARs for an unseen CT scan is 9.5 s. CONCLUSION: Experiments on clinical datasets of H&N patients demonstrated the effectiveness of the proposed deep neural Network Segmentation method for multi-organ Segmentation on volumetric CT scans. The accuracy and robustness of the Segmentation were further increased by incorporating shape priors using SMR. The proposed method showed competitive performance and took shorter time to segment multiple organs in comparison to state of the art methods.
Raditya Arief - One of the best experts on this subject based on the ideXlab platform.
-
mitigating cyberattack related domino effects in process plants via ics Segmentation
Workshop on Information Security Applications, 2020Co-Authors: Raditya Arief, Nima Khakzad, Wolter PietersAbstract:Abstract Domino effects are high-impact phenomena that have caused catastrophic damage to several chemical and process plants around the world through secondary incidents caused by primary ones. With the increasing trend of cyberattacks targeting critical infrastructures, there is a concern that such cyberattacks may trigger domino effects, by manipulating industrial control systems in such a way that the physical consequences are likely to escalate. In this study, we have demonstrated that via Network Segmentation of industrial control systems, the plant robustness against cyberattack-related domino effects can be improved. To this end, a risk-based decision-making methodology is developed based on Bayesian Network and graph theory to investigate and evaluate the robustness of Segmentation alternatives. The application of the methodology to an illustrative case study shows the efficacy of the approach as a viable cyber risk mitigation measure in chemical and process plants.
-
cyberattack related cascading effects mitigation a risk based approach for ics Network Segmentation design in chemical plants
2018Co-Authors: Raditya AriefAbstract:Cascading effects are high-impact, low-probability phenomena that have caused catastrophic impacts in various chemical and process plants around the world. With the increasing trend of cyberattacks targeting critical infrastructures, there is a concern that accidents caused by cyberattacks may trigger cascading effects in these facilities. In this study, we have demonstrated that the implementation of Network Segmentation to ICS Networks to improve its robustness against the risk of cyberattack-related cascading effects. A risk-based methodology is developed to investigate and evaluate the robustness of design alternatives. The risk-based methodology also incorporates a risk assessment method based on Bayesian Networks for cascading effects modeling. Further, this thesis also presents some design guidelines for developing robust Networks Segmentation designs, which includes the application of a graph-theoretic approach that enables early identification of the severity of cascading effects in Network design alternatives. The risk-based methodology and the design guidelines are applied to a case study in which the efficacy of the approaches are demonstrated. The outcome of the risk-based methodology offers an alternative risk mitigation technique that can be considered in the risk management process.
K Sheng - One of the best experts on this subject based on the ideXlab platform.
-
fully automatic multi organ Segmentation for head and neck cancer radiotherapy using shape representation model constrained fully convolutional neural Networks
Medical Physics, 2018Co-Authors: Nuo Tong, Shuyuan Yang, Dan Ruan, K ShengAbstract:PURPOSE: Intensity modulated radiation therapy (IMRT) is commonly employed for treating head and neck (HN (b) the pre-trained SRM with fixed parameters were used to constrain the FCNN training. The combined Segmentation Network was then used to delineate nine OARs including the brainstem, optic chiasm, mandible, optical nerves, parotids, and submandibular glands on unseen H&N CT images. Twenty-two and 10 H&N CT scans provided by the Public Domain Database for Computational Anatomy (PDDCA) were utilized for training and validation, respectively. Dice similarity coefficient (DSC), positive predictive value (PPV), sensitivity (SEN), average surface distance (ASD), and 95% maximum surface distance (95%SD) were calculated to quantitatively evaluate the Segmentation accuracy of the proposed method. The proposed method was compared with an active appearance model that won the 2015 MICCAI H&N Segmentation Grand Challenge based on the same dataset, an atlas method and a deep learning method based on different patient datasets. RESULTS: An average DSC = 0.870 (brainstem), DSC = 0.583 (optic chiasm), DSC = 0.937 (mandible), DSC = 0.653 (left optic nerve), DSC = 0.689 (right optic nerve), DSC = 0.835 (left parotid), DSC = 0.832 (right parotid), DSC = 0.755 (left submandibular), and DSC = 0.813 (right submandibular) were achieved. The Segmentation results are consistently superior to the results of atlas and statistical shape based methods as well as a patch-wise convolutional neural Network method. Once the Networks are trained off-line, the average time to segment all 9 OARs for an unseen CT scan is 9.5 s. CONCLUSION: Experiments on clinical datasets of H&N patients demonstrated the effectiveness of the proposed deep neural Network Segmentation method for multi-organ Segmentation on volumetric CT scans. The accuracy and robustness of the Segmentation were further increased by incorporating shape priors using SMR. The proposed method showed competitive performance and took shorter time to segment multiple organs in comparison to state of the art methods.
Unamay Oreilly - One of the best experts on this subject based on the ideXlab platform.
-
adversarial co evolution of attack and defense in a segmented computer Network environment
Genetic and Evolutionary Computation Conference, 2018Co-Authors: Erik Hemberg, Neal Wagner, Joseph R Zipkin, Richard Skowyra, Unamay OreillyAbstract:In computer security, guidance is slim on how to prioritize or configure the many available defensive measures, when guidance is available at all. We show how a competitive co-evolutionary algorithm framework can identify defensive configurations that are effective against a range of attackers. We consider Network Segmentation, a widely recommended defensive strategy, deployed against the threat of serial Network security attacks that delay the mission of the Network's operator. We employ a simulation model to investigate the effectiveness over time of different defensive strategies against different attack strategies. For a set of four Network topologies, we generate strong availability attack patterns that were not identified a priori. Then, by combining the simulation with a co-evolutionary algorithm to explore the adversaries' action spaces, we identify effective configurations that minimize mission delay when facing the attacks. The novel application of co-evolutionary computation to enterprise Network security represents a step toward course-of-action determination that is robust to responses by intelligent adversaries.1
Neal Wagner - One of the best experts on this subject based on the ideXlab platform.
-
adversarial co evolution of attack and defense in a segmented computer Network environment
Genetic and Evolutionary Computation Conference, 2018Co-Authors: Erik Hemberg, Neal Wagner, Joseph R Zipkin, Richard Skowyra, Unamay OreillyAbstract:In computer security, guidance is slim on how to prioritize or configure the many available defensive measures, when guidance is available at all. We show how a competitive co-evolutionary algorithm framework can identify defensive configurations that are effective against a range of attackers. We consider Network Segmentation, a widely recommended defensive strategy, deployed against the threat of serial Network security attacks that delay the mission of the Network's operator. We employ a simulation model to investigate the effectiveness over time of different defensive strategies against different attack strategies. For a set of four Network topologies, we generate strong availability attack patterns that were not identified a priori. Then, by combining the simulation with a co-evolutionary algorithm to explore the adversaries' action spaces, we identify effective configurations that minimize mission delay when facing the attacks. The novel application of co-evolutionary computation to enterprise Network security represents a step toward course-of-action determination that is robust to responses by intelligent adversaries.1
-
capturing the security effects of Network Segmentation via a continuous time markov chain model
Annual Simulation Symposium, 2017Co-Authors: Neal Wagner, Cem şafak şahin, Jaime Pena, James Riordan, Sebastian NeumayerAbstract:Segmentation or compartmentalization of a computer Network is a commonly used defensive mitigation against cyber attack. Its goal is to limit the damage an attacker can cause by partitioning a Network into sections or enclaves and restricting communications between them. While this technique is widely advocated as critical to the security of a Network, no clear guidance currently exists on how to appropriately implement it. Additionally, the cost of testing candidate Segmentation architectures on a live Network or a cyber test environment is prohibitively expensive. This study examines an alternative method for evaluating Segmentation architectures utilizing a continuous-time Markov chain to model changes in Network state based on relevant Network parameters such as vulnerability arrival rate, patch rate, etc. The model is realized by an event-based Network simulation and demonstrated via a case study that evaluates a range of candidate architectures.