Feedback

Computational complexity evaluation of ann algorithms for image steganalysis

Documento informativo
International Journal of Latest Trends in Engineering and Technology (IJLTET) Computational Complexity Evaluation of ANN Algorithms for Image Steganalysis Dr.P.Sujatha Assistant Professor, Department of Computer Science Vels University, Chennai, India. Dr.S.Purushothaman Professor, PET Engineering college, Vallioor – 627 117 Tirunelveli Dt., India, P.Rajeswari Research Scholar, Mother Therasa University Kodaikanal, India, Abstract - The major growth of information technology is based on the way how the security measures are implemented. Steganography is a technique that implements high level security by hiding a message in a multimedia object such as image. Steganalysis is the way of detecting such hidden messages. In order to detect the presence of hidden message, artificial neural network algorithms such as Back Propagation and Radial Basis Function are used. This paper performs the computational complexity evaluation of these two algorithms. Keywords - Covert communication, Steganography, Steganalysis, ANN, Back Propagation, Radial Basis Function I. INTRODUCTION In today’s digital age, there are more chances for altering the information represented by an image without leaving any traces of tampering. Many areas such as forensics investigation, surveillance systems, criminal investigation, medical imaging, journalism and intelligence services need reliability while transferring the information in the form of an image. Planning is the crucial part and the information planning is passed to others through covert communication in order to hide from government and other people. The effective medium of hidden communication is achieved by steganography. An article (Jack [11]) ensured that terrorists used steganography for secret communication during 11th September 2001 attack. Politicians use steganography communication to express their political thoughts that are more sensitive to the world. The Government can take action on any politician who involves in sensitive issue like decreasing the economical growth of the country. The ease of Internet helps in both good and bad usage. Downloading various tools for steganography becomes a challenging task for government to trace the law breakers. The majorities of documents used in publishing industry were digital documents with foreground (black) and background (White) binary values. Multiple documents are manipulated everyday with binary values. Those documents are scanned and used as a medium of steganography. Variety of data embedding algorithms and variety of images that makes the steganography a toughest mission for researchers to develop a powerful technique for steganalysis. II. IMAGE STEGANALYSIS The counter-technique of image steganography is known as image steganalysis. Steganalysis begins by identifying the artifacts that exist in the suspectable file which is a result of message embedding. The goal is not to advocate the removal or disabling of valid hidden information such as copyrights, but to point out approaches that are vulnerable and may be exploited to investigate illicit hidden information (Anderson et al. [1]; Johnson et al. [2]; Neil et al. [7]; Rajarathnam et al. [9]). Attacks and analysis on hidden information may take several forms like detecting, extracting, and disabling or destroying hidden information (Westfeld et al. [3]). An attacker may also embed counter-information over the existing hidden information. These approaches vary depending upon the methods used to embed the information into the cover media. Vol. 3 Issue 3 January 2014 229 ISSN: 2278-621X International Journal of Latest Trends in Engineering and Technology (IJLTET) A. Steganalysis Methods Based on the way of detecting the presence of hidden message, steganalysis methods are divided as follows. i) Statistical steganalysis a) Spatial domain b) Transform domain ii) Feature based steganalysis Statistical Steganalysis: Existence of the hidden message is detected using statistical analysis that is done with the pixels. It is further classified as spatial domain steganalysis and transform domain steganalysis. In spatial domain, the pair of pixels is considered and the difference between them is calculated. In transform domain, frequency counts of coefficients are calculated and then histogram analysis is performed. Feature based steganalysis: The features of the image are used to detect hidden message in an image. They can also be used to train classifiers. III. RELATED WORKS Fridrich [5] developed a steganalytic technique that detects LSB embedding in color and grayscale images. They analyze the capacity for embedding lossless data in LSBs. Randomizing the LSBs decreases this capacity. To examine an image, they define Regular groups (R) and Singular groups (S) of pixels depending upon some properties. With the help of relative frequencies of these groups in the given image along with an image obtained from the original image with LSBs flipped and with an image obtained by randomizing LSBs of the original image, they try to predict the levels of embedding. Fridrich [8] proposed Pairs analysis method. This approach is well suited for the embedding archetype that randomly embeds messages in LSBs of indices to palette colors of palette image. Westfeld [4] used visual attacks to detect the steganography by making use of the ability of human eyes to inspect the images for the corruption caused by the embedding. Martin [6] attempts to directly use the notion of the naturalness of images to detect hidden data. Though they found that data hidden certainly caused shifts from the natural set, knowledge of the specific data hiding scheme provides far better detection performance. IV. PROPOSED ALGORITHM Apart from all modern sciences and technologies, Artificial Neural Network (ANN) plays a vital role in capturing and representing both linear and non-linear relationships. ANN is an intelligent system which helps to enable machines to solve problems like human by extracting and storing the knowledge. To incorporate intelligent method for steganalysis, this paper focuses ANN to overcome the drawbacks of the conventional steganalysis methods. The proposed methods are, x Back Propagation Algorithm (BPA) x Radial Basis Function (RBF) A. Implementation of BPA: The BPA uses the steepest-descent method to reach a global minimum. The number of layers and number of nodes in the hidden layers are decided. The connections between nodes are initialized with random weights. A pattern from the training set is presented in the input layer of the network and the error at the output layer is calculated. The error is propagated backward towards the input layer and the weights are updated. At the end of each iteration, test patterns are presented to ANN, and the prediction performance of ANN is evaluated. Further training of ANN is continued till the desired prediction performance is reached. Vol. 3 Issue 3 January 2014 230 ISSN: 2278-621X International Journal of Latest Trends in Engineering and Technology (IJLTET) Figure 1. Detection of location of message by BPA In Figure 1, “ƔOriginal” refers to the actual information of the image. “ Detected” information indicates that the suspect image is a stegnographic image. B. Implementation of RBF: (YHU\IXQFWLRQFDQEHXQLTXHO\LGHQWLILHGE\LWVLQKHUHQWSURSHUWLHVDQGWKLVPDNHVDIRUPRIɎVXLWDEOHLQ approximation to one problem or a particular class of problems. The selection of position and the number of centers is similar to problems choosing the number and initial values of the weights in a multilayer perceptron (MLP). A best approximation can be produced when optimal number of centers is identified. Neither very few nor many centers should be chosen, since this may lead to poor approximation. It is very important to maintain equilibrium between the number of centers and the amount of training data. Figure 2. Detection of Message location using RBF In the above figure, the detected information is represented by ‘’. The pixels in cover image are represented by ‘Ɣ’. IV. COMPARISON OF PERFORMANCES A. Computational Complexity The computational complexity of an algorithm is defined as the number of arithmetic operations required for training the proposed algorithm. The performance comparison of modified ANN algorithm (Vu et al. [10]) is studied for. Empirical formulae for computational effort are presented for BPA and RBF in Table 1. Table 1. Computational Complexity Evaluation Algorithm Formula for evaluating the number of arthmetic computation BPA Forward computational effort in BPA for one pattern is given by L1 2 ¦n i 1 (n i  1) (1.1) i 1 Reverse computational effort in BPA for one pattern is given by L1 2 9n L  7¦ n i n i1  i 1 ¦ (4n TCE for BPA= {(ite) a o } n p Vol. 3 Issue 3 January 2014 i  5)n i1 (1.2) i L1 231 (1.3) ISSN: 2278-621X International Journal of Latest Trends in Engineering and Technology (IJLTET) TCE for RBF = {2n c 2 +inv(n c 2) n c 2 } n p (1.4) RBF where: TCE Total computational effort, n c is number of centers+1(bias) , n p is number of training patterns, inv is inbuilt function inverse of a matrix, ite is the number of iteration a o is {Forward computation in BPA + Reverse computation in BPA} L is the total number of layers including the input layer, ‘i’ is the layer number, and (n i ) is the number of nodes in the ith layer B. Computational effort comparison: Results of the computational effort comparison of proposed algorithms are given in Table 2. By using equations given in Table 1, the computational effort for each algorithm has been calculated and presented, in order to compare them with regard to the total number of computational effort required by each algorithm. 6 0.000811216 111 2 RBF 2(2) 1 NA 32 MSE Computational effort 2(2) iterations Number of BPA Algorithm 1 S.No. No. of nodes in input layer (Hidden layer) Table 2. Computational effort comparison for proposed algorithms. The network trained with transformed vector requires the least computational effort. RBF algorithm needs less computational effort than BPA. COMPUTATIONAL EFFORT COMPUTATIONAL COMPLEXITY EVALUATION 120 111 100 80 60 40 20 0 BPA RBF 32 6 1 NUMBER OF ITERATIONS . V. CONCLUSION This paper proposed the most popular supervised artificial neural network algorithms such as BPA and RBF for detecting the presence of the hidden message. The computational effort is also compared for the proposed algorithms. The comparison shows that RBF needs less computational effort than BPA. REFERENCES [1] [2] [3] [4] Anderson, R.J., and Petitcolas, F.A.P., 1998, “On the limits of steganography”, IEEE Journal on Selected Areas in Communications, Vol. 16, No. 4, pp. 474-484. Johnson, N., and Sklansky, J., 1998, “Exploring steganography: seeing the unseen”, IEEE Computer, pp. 26-34. Westfeld, A., and Pfitzmann, A., 1999, “Attacks on steganographic systems”, Lecture notes in computer science, proceedings of the 3rd International Workshop on Information Hiding, Vol. 1768, pp. 61-76. Westfeld, A., and Pfitzmann, A., 2000, “Attacks on Steganographic Systems”, Lecture Notes in Computer Science, Springer-Verlag, Berlin, Vol. 1768, pp. 61-75. Vol. 3 Issue 3 January 2014 232 ISSN: 2278-621X International Journal of Latest Trends in Engineering and Technology (IJLTET) [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] Fridrich, J., Goljan, M., and Du R., 2001, “Reliable Detection of LSB Steganography in Color and Gray-Scale Images”, Magazine of IEEE Multimedia Special Issue on Security, pp. 22-28. Martin, A., Sapiro G., and Seroussi G., 2004, “Is image steganography natural?”, Information Theory Research Group, HP Laboratories Palo Alto, HPL-39(R.1). Neil Provos, and Peter Honeyman, 2003, “Hide and Seek: An Introduction to Steganography”, IEEE Security & Privacy, Vol. 1, No. 3, pp. 32-44. Fridrich, J., and Goljan, M., 2003, “Digital image steganography using stochastic modulation”, Proceedings of IST/SPIE’s 15th Annual Symposium on Electronic Imaging Science and Technology, San Jose, CA. Rajarathnam, Chandramouli, Mehdi Kharrazi, and Nasir Memon, 2004, “Image Steganography and Steganalysis: Concepts and Practice”, Lecture Notes in Computer Science, International Workshop on Digital Watermarking, Korea, Vol. 2939, pp. 35-49. Vu Dao, N.P., and Rao Vemuri, 2002, “A Performance Comparison of Different Back Propagation Neural Networks Methods in Computer Network Intrusion Detection”, Differential Equations and Dynamical Systems, Vol. 10, No. 1, pp. 201-214. Jack Kelley., 2001, “Terror Groups Hide Behind Web Encryption”, USA Today. Natarajan, V., and Anitha, R., 2012, “Blind Image steganalysis Based on Contourlet Transform”, International journal on Cryptography & Information Security, Vol. 2, No. 3, pp. 77-87. Chin-Chen Chang, Lin, CY., and Fan, YH., 2008, “Lossless data hiding for color images based on block truncation coding”, Pattern Recognition, Vol. 41, No. 7, pp. 2347–2357. Yun Shi, Q., Guorong Xuan, Chengyun Yang, Jianjiong Gao, Zhenping Zhang, Peiqi Chai, Dekun Zou, Chunhua Chen, and Wen Chen, 2005, “Effective steganalysis Based on Statistical Moments of Wavelet Characteristic Function”, IEEE International Conference on Information Technology: Coding and Computing, Vol. 1, pp. 768-773. Yong Wang, JiuFen Liu, and WeiMing Zhang, 2009, “Blind JPEG steganalysis Based on Correlations of DCT Coefficients in Multidirections and Calibrations”, International Conference on Multimedia Information Networking and Security, MINES'09, Vol. 1, pp. 495499. Xiao Yi Yu, and Aiming Wang, 2009, “Steganalysis Based on Bayesion Network and Genetic Algorithm”, 2nd International Congress on Image and Signal Processing, CISP'09, pp. 1-4. Declan Mc Cullah., 2001, “Secret Messages Come in Wavs”, Wired News. Daniel Lerch-+RVWDORWDQG'DYLG0HJÕDV“Steganalytic Methods for the Detection of Histogram Shifting Data Hiding Schemes”, Proceedings of Reunión Española Cryptology and Information Security (RECSI). Vol. 3 Issue 3 January 2014 233 ISSN: 2278-621X active. HIV-infected patients displayed a redistribution of body fat as the percentage of fat on the limb were lower and the percentage of fat in the trunk was higher compared to control subjects. The patients also had disturbances in their lipid metabolism as fasting triglycerides and total-cholesterol levels were higher, and HDL-cholesterol level was lower. In addition, compared to controls, the HIV-infected patients were characterized by peripheral insulin resistance as whole-body insulin-stimulated glucose uptake (Rd) and incremental glucose uptake (Rd basal – Rd clamp) were decreased. Furthermore, the HIV subjects had higher basal endogenous glucose production, lower insulin-mediated suppression of endogenous glucose production (Ra), but no difference in the incremental suppression of endogenous glucose production during the clamp (Ra basal – Ra clamp) as compared to control subjects. Plasma FGF-21 and FGF-21mRNA in muscle Plasma FGF-21 was elevated in the HIV-group compared to healthy controls (70.4656.8 pg/ml vs 109.1671.8 pg/ml, respectively) (P = 0.04) (Figure 1A). FGF-21 mRNA expression in skeletal muscle was increased 8fold in patients with HIV relative to healthy individuals (P,0.0001, parametric statistics, and p = 0.0002 for non-para- metric statistics) (Figure 1B). The association between plasma FGF-21 and muscle FGF-21 did not reach statistical significance (r = 0.32; p = 0.056). Relationships between FGF-21 mRNA, plasma FGF-21 and insulin resistance Muscle FGF-21 mRNA correlated positively to all markers of insulin resistance: fasting insulin (r = 0.57, p = 0.0008), homeostasis model assessment (HOMA) score (r = 0.55, p = 0.001), and insulinAUC (r = 0.38, p = 0.02) (Fig. 1 C–F). To test whether the association between insulin and FGF-21 mRNA in muscle reflect an association with peripheral or hepatic insulin resistance, we performed a euglycemic-hyperinsulinemic clamp with stable isotopes. We found that muscle FGF-21 mRNA was negatively associated with the insulin-mediated glucose-uptake (Figure 1G), but not with hepatic insulin resistance (data not shown). We did not find any association between plasma FGF-21 and parameters of insulin resistance (fasting insulin, plasma glucose, HOMA-IR, glucose AUC, Insulin AUC, Ra or Rd). However, when investigating by group, plasma FGF-21 correlated positive with insulin stimulated glucose-uptake in healthy subjects but not in the HIV patients (r = 0.51, p = 0.049) (data not shown). Previous studies have demonstrated that insulin resistance in patients with HIV-LD are associated with decreased insulinstimulated glycogen synthase (GS) activity [24]. Therefore, we measured GS activity. The GS fractional velocity of % total GS activity was lower in HIV patients compared with healthy controls (2861.3; 3561.6, respectively) in the basal stage. GS fractional velocity is known to correlate positively with the glucose Rd [25]. As high FGF-21 mRNA in muscle is associated with low rate of disappearance of glucose, this could be linked to low GS fractional velocity in muscle. In accordance with this hypothesis, we found that high levels of FGF-21 mRNA in muscle were associated with decreased GS fractional velocity in muscle (Fig. 1H). Relationships between muscle FGF-21 mRNA, and fat distribution and lipids We found a strong negative association between muscle FGF-21 mRNA and the amount of subcutaneous fat (limb fat mass) (r = 20.46; p = 0.0038) and positive association with trunk-limbfat-ratio (r = 0.51; p = 0.001) and triglycerides (r = 0.56; p = 0.0003). FGF-21 mRNA was not associated with total fat mass or total trunk fat (data not shown). Discussion The novelty and the major findings of our study is that we demonstrate for the first time that FGF-21 mRNA expression is increased in skeletal muscle in patients with HIV-LD compared to healthy age-matched men and that muscle FGF-21 mRNA correlates negatively with the rate of insulin-stimulated glucose disappearance (primarily reflecting muscle). Furthermore, increased FGF-21 mRNA expression in muscle is associated with decreased limb fat mass, increased waist-to-hip ratio and increased triglycerides. Only three studies are published on expression of FGF-21 in muscle in humans [10,12,26] and very little is known about the function of muscle FGF-21. Our result is in agreement with a previous study, in which FGF-21 mRNA was found to be increased in muscle from subjects with type 2 diabetes and the expression was increased by hyperinsulinemia [10]. However, this study did not distinguish between hepatic and peripheral insulin sensitivity as we do in the present study. Vienberg et al. [12] also PLOS ONE | www.plosone.org 3 March 2013 | Volume 8 | Issue 3 | e55632 Muscle FGF-21,Insulin Resistance and Lipodystrophy Table 1. Baseline characteristics of patients and healthy controls. Age (years) Duration of HIV infection (years) Duration of antiretroviral therapy (years) CD4+ cell (cells/ml) LogHIV-RNA (copies/ml) Antiretroviral use NNRTI-based HAART, PI-based HAART, NNRTI-,PI-based HAART regime, No. Current Thymidine-NRTI use, No. (%) Current PI use, No. (%) Current NNRTI use, No. (%) Physical activity parameters VO2max (LO2/min) Body composition Body-mass index (kg/m2) Weight (kg) Waist (cm) Waist-to.hip ratio Fat mass (kg) Trunk fat mass (kg) Trunk fat percentage (%) Limb fat mass (kg) Limb fat percentage (%) Trunk-to-limb fat ratio Lean mass (kg) Metabolic parameters Total-cholesterol (mmol/L) HDL-C (mmol/L) LDL-C (mmol/L) Triglycerides (mmol/L) Glucose (mmol/L) Insulin (pmol/L) HOMA-IR Insulin sensitivity Rate of appearance (mmol glucose/kg/min) Basal Clamp Delta{ Rate of disappearance (mmol glucose/kg/min) Basal Clamp Delta{ Glucose tolerance Glucose area under the curve (mmol/Lmin) Insulin area under the curve (pmol/Lmin) HIV patients (n = 23) 47.9 (9.5) 15.6 (9.6) 10.3 (4.3) 558 (208) 1.33 (0.12) 12/14/2 11 (47.8) 13 (56.7) 11 (47.8) 2.3 (0.5) 23.7 (2.9) 73.6 (11.2) 93.6 (6.4) 1.01 (0.04) 13.8 (5.3) 9.8 (3.9) 71.2 (6.2) **** 3.5 (1.6)**** 25.1 (6.1) **** 3.09 (1.17)* 57.0 (6.8) 5.5 (0.9)** 1.23 (0.52)* 3.7 (0.9) 2.55(1.43)**** 5.4 (0.6) 52 (25)**** 2.2 (1.4)**** 14.2 (0.49)*** 6.4 (1.8)** 7.8 (2.1) 14.2 (0.49)*** 40.2 (9.9)** 26.02 (10.1)** 826 (200)* 52360 (31017)** Healthy controls (n = 15) 47.5 (6.1) 2.5 (0.6) 23.7 (1.9) 76.9 (7.4) 90 (5.7) 0.94 (0.03) 15.7 (4.4) 8.9 (3.0) 56.1(5.2) 6.2 (1.5) 40.2 (4.9) 1.4 (0.29) 58.2 (5.2) 4.63 (0.64) 1.51 (0.32) 3.3 (0.6) 0.76 (0.24) 5.2 (0.3) 25 (8.9) 0.99 (0.37) 11.8 (2.0) 4.0 (2.5) 7.8 (1.9) 11.8 (2.0) 48.6 (8.4) 36.81 (7.14) 670 (126) 23505 (10598) Data are presented as mean (SD). {Delta, differences between clamp and basal values. HAART, highly active antiretroviral therapy; PI, protease inhibitor; NRTI, nucleoside reverse transcriptase inhibitor; NNRTI, non- nucleoside reverse transcriptase inhibitor. HOMA-IR, homeostatic model assessment for insulin resistance, Rate of appearance and disappearance, Rate of appearance and disappearance of glucose during a euglycemic-hyperinsulinemic clamp performed in both HIV patients and healthy controls. *P,0.05; **P,0.01; ***P,0.001, ****P,0.0001 by t-test. doi:10.1371/journal.pone.0055632.t001 PLOS ONE | www.plosone.org 4 March 2013 | Volume 8 | Issue 3 | e55632 Muscle FGF-21,Insulin Resistance and Lipodystrophy Figure 1. FGF-21 mRNA are increased in muscle from subjects with HIV-lipodystrophy and correlates to several measurement of insulin resistance. (A) Fasting plasma levels of fibroblast growth factor (FGF) 21 are increased 2-fold in HIV subjects with lipodystrophy compared to healthy men; (B) mRNA expression of FGF-21 are increased 8-fold in muscle biopsies from HIV subjects with lipodystrophy compared to healthy men; (C–F) Plots of FGF-21 mRNA in muscle versus several measurements of insulin resistance: FGF-21 mRNA in muscle are positively correlated to fasting insulin (C), HOMA-IR (D), Area under the curve for insulin during an oral glucose tolerance test (E), Area under the curve for C-peptide during Nan oral glucose tolerance test (F), and negatively correlated to the incremental rate of disappearance of glucose (G), and fractionel velocity of glycogen synthesis (H) in healthy (e) and HIV subjects with lipodystrophy ( ). In the dot plots data for each subjects are given and the line represent means. * P,0.05 and ***P,0.001 for healthy vs HIV-lipodystrophy patients. For plots, linear regression lines, correlations coefficient, and significance levels are given for all subjects. doi:10.1371/journal.pone.0055632.g001 PLOS ONE | www.plosone.org 5 March 2013 | Volume 8 | Issue 3 | e55632 Muscle FGF-21,Insulin Resistance and Lipodystrophy find that FGF-21 mRNA was expressed in muscle, but they find no activation of muscle FGF-21 mRNA by a short term high fat overfeeding. In the study by Mashili et al. [26], FGF-21 mRNA was expressed in skeletal muscle, but in the aquatic environment for long-term survival and for transmission to arthropod and mammalian hosts. References 1. Ray K, Marteyn B, Sansonetti PJ, Tang CM. Life on the inside: the intracellular lifestyle of cytosolic bacteria. Nature reviews. 2009; 7(5):333–40. doi: 10.1038/nrmicro2112 PMID: 19369949 2. Kingry LC, Petersen JM. Comparative review of Francisella tularensis and Francisella novicida. Frontiers in cellular and infection microbiology. 2014; 4:35. doi: 10.3389/fcimb.2014.00035 PMID: 24660164 3. Sjodin A, Svensson K, Ohrman C, Ahlinder J, Lindgren P, Duodu S, et al. Genome characterisation of the genus Francisella reveals insight into similar evolutionary paths in pathogens of mammals and fish. BMC genomics. 2012; 13:268. doi: 10.1186/1471-2164-13-268 PMID: 22727144 4. Ellis J, Oyston PC, Green M, Titball RW. Tularemia. Clinical microbiology reviews. 2002; 15(4):631–46. PMID: 12364373 5. Gurcan S. Epidemiology of tularemia. Balkan medical journal. 2014; 31(1):3–10. doi: 10.5152/ balkanmedj.2014.13117 PMID: 25207161 6. Petersen JM, Mead PS, Schriefer ME. Francisella tularensis: an arthropod-borne pathogen. Veterinary research. 2009; 40(2):7. doi: 10.1051/vetres:2008045 PMID: 18950590 7. Lundstrom JO, Andersson AC, Backman S, Schafer ML, Forsman M, Thelaus J. Transstadial transmission of Francisella tularensis holarctica in mosquitoes, Sweden. Emerging infectious diseases. 2011; 17(5):794–9. doi: 10.3201/eid1705.100426 PMID: 21529386 8. van Hoek ML. Biofilms: an advancement in our understanding of Francisella species. Virulence. 2013; 4(8):833–46. doi: 10.4161/viru.27023 PMID: 24225421 9. Ryden P, Sjostedt A, Johansson A. Effects of climate change on tularaemia disease activity in Sweden. Global health action. 2009; 2. PLOS Pathogens | DOI:10.1371/journal.ppat.1005208 December 3, 2015 6/8 10. Barker JR, Chong A, Wehrly TD, Yu JJ, Rodriguez SA, Liu J, et al. The Francisella tularensis pathogenicity island encodes a secretion system that is required for phagosome escape and virulence. Molecular microbiology. 2009; 74(6):1459–70. PMID: 20054881 11. de Bruin OM, Duplantis BN, Ludu JS, Hare RF, Nix EB, Schmerk CL, et al. The biochemical properties of the Francisella pathogenicity island (FPI)-encoded proteins IglA, IglB, IglC, PdpB and DotU suggest roles in type VI secretion. Microbiology (Reading, England). 2011; 157(Pt 12):3483–91. 12. Filloux A, Hachani A, Bleves S. The bacterial type VI secretion machine: yet another player for protein transport across membranes. Microbiology (Reading, England). 2008; 154(Pt 6):1570–83. 13. Nguyen JQ, Gilley RP, Zogaj X, Rodriguez SA, Klose KE. Lipidation of the FPI protein IglE contributes to Francisella tularensis ssp. novicida intramacrophage replication and virulence. Pathogens and disease. 2014; 72(1):10–8. doi: 10.1111/2049-632X.12167 PMID: 24616435 14. Hood RD, Singh P, Hsu F, Guvener T, Carl MA, Trinidad RR, et al. A type VI secretion system of Pseudomonas aeruginosa targets a toxin to bacteria. Cell host & microbe. 2010; 7(1):25–37. 15. Mougous JD, Cuff ME, Raunser S, Shen A, Zhou M, Gifford CA, et al. A virulence locus of Pseudomonas aeruginosa encodes a protein secretion apparatus. Science (New York, NY. 2006; 312 (5779):1526–30. 16. Pukatzki S, Ma AT, Revel AT, Sturtevant D, Mekalanos JJ. Type VI secretion system translocates a phage tail spike-like protein into target cells where it cross-links actin. Proceedings of the National Academy of Sciences of the United States of America. 2007; 104(39):15508–13. PMID: 17873062 17. Santic M, Molmeret M, Klose KE, Jones S, Kwaik YA. The Francisella tularensis pathogenicity island protein IglC and its regulator MglA are essential for modulating phagosome biogenesis and subsequent bacterial escape into the cytoplasm. Cellular microbiology. 2005; 7(7):969–79. PMID: 15953029 18. de Bruin OM, Ludu JS, Nano FE. The Francisella pathogenicity island protein IglA localizes to the bacterial cytoplasm and is needed for intracellular growth. BMC microbiology. 2007; 7:1. PMID: 17233889 19. Llewellyn AC, Jones CL, Napier BA, Bina JE, Weiss DS. Macrophage replication screen identifies a novel Francisella hydroperoxide resistance protein involved in virulence. PloS one. 2011; 6(9):e24201. doi: 10.1371/journal.pone.0024201 PMID: 21915295 20. Clemens DL, Lee BY, Horwitz MA. Virulent and avirulent strains of Francisella tularensis prevent acidification and maturation of their phagosomes and escape into the cytoplasm in human macrophages. Infection and immunity. 2004; 72(6):3204–17. PMID: 15155622 21. Moreau GB, Mann BJ. Adherence and uptake of Francisella into host cells. Virulence. 2013; 4(8):826– 32. doi: 10.4161/viru.25629 PMID: 23921460 22. Tamilselvam B, Daefler S. Francisella targets cholesterol-rich host cell membrane domains for entry into macrophages. J Immunol. 2008; 180(12):8262–71. PMID: 18523292 23. Chong A, Wehrly TD, Nair V, Fischer ER, Barker JR, Klose KE, et al. The early phagosomal stage of Francisella tularensis determines optimal phagosomal escape and Francisella pathogenicity island protein expression. Infection and immunity. 2008; 76(12):5488–99. doi: 10.1128/IAI.00682-08 PMID: 18852245 24. Celli J, Zahrt TC. Mechanisms of Francisella tularensis intracellular pathogenesis. Cold Spring Harbor perspectives in medicine. 2013; 3(4):a010314. doi: 10.1101/cshperspect.a010314 PMID: 23545572 25. Santic M, Asare R, Skrobonja I, Jones S, Abu Kwaik Y. Acquisition of the vacuolar ATPase proton pump and phagosome acidification are essential for escape of Francisella tularensis into the macrophage cytosol. Infection and immunity. 2008; 76(6):2671–7. doi: 10.1128/IAI.00185-08 PMID: 18390995 26. Jones CL, Napier BA, Sampson TR, Llewellyn AC, Schroeder MR, Weiss DS. Subversion of host recognition and defense systems by Francisella spp. Microbiol Mol Biol Rev. 2012; 76(2):383–404. doi: 10. 1128/MMBR.05027-11 PMID: 22688817 27. Pierini R, Juruj C, Perret M, Jones CL, Mangeot P, Weiss DS, et al. AIM2/ASC triggers caspase-8dependent apoptosis in Francisella-infected caspase-1-deficient macrophages. Cell death and differentiation. 2012; 19(10):1709–21. doi: 10.1038/cdd.2012.51 PMID: 22555457 28. Asare R, Kwaik YA. Exploitation of host cell biology and evasion of immunity by francisella tularensis. Frontiers in microbiology. 2011; 1:145. doi: 10.3389/fmicb.2010.00145 PMID: 21687747 29. Keim P, Johansson A, Wagner DM. Molecular epidemiology, evolution, and ecology of Francisella. Annals of the New York Academy of Sciences. 2007; 1105:30–66. PMID: 17435120 30. Asare R, Akimana C, Jones S, Abu Kwaik Y. Molecular bases of proliferation of Francisella tularensis in arthropod vectors. Environmental microbiology. 2010; 12(9):2587–612. doi: 10.1111/j.1462-2920. 2010.02230.x PMID: 20482589 PLOS Pathogens | DOI:10.1371/journal.ppat.1005208 December 3, 2015 7/8 31. Vonkavaara M, Telepnev MV, Ryden P, Sjostedt A, Stoven S. Drosophila melanogaster as a model for elucidating the pathogenicity of Francisella tularensis. Cellular microbiology. 2008; 10(6):1327–38. doi: 10.1111/j.1462-5822.2008.01129.x PMID: 18248629 32. Santic M, Akimana C, Asare R, Kouokam JC, Atay S, Kwaik YA. Intracellular fate of Francisella tularensis within arthropod-derived cells. Environmental microbiology. 2009; 11(6):1473–81. doi: 10.1111/j. 1462-2920.2009.01875.x PMID: 19220402 33. Read A, Vogl SJ, Hueffer K, Gallagher LA, Happ GM. Francisella genes required for replication in mosquito cells. Journal of medical entomology. 2008; 45(6):1108–16. PMID: 19058636 34. Ahlund MK, Ryden P, Sjostedt A, Stoven S. Directed screen of Francisella novicida virulence determinants using Drosophila melanogaster. Infection and immunity. 2010; 78(7):3118–28. doi: 10.1128/IAI. 00146-10 PMID: 20479082 35. Abd H, Johansson T, Golovliov I, Sandstrom G, Forsman M. Survival and growth of Francisella tularensis in Acanthamoeba castellanii. Applied and environmental microbiology. 2003; 69(1):600–6. PMID: 12514047 36. Santic M, Ozanic M, Semic V, Pavokovic G, Mrvcic V, Kwaik YA. Intra-Vacuolar Proliferation of F. Novicida within H. Vermiformis. Frontiers in microbiology. 2011; 2:78. doi: 10.3389/fmicb.2011.00078 PMID: 21747796 37. El-Etr SH, Margolis JJ, Monack D, Robison RA, Cohen M, Moore E, et al. Francisella tularensis type A strains cause the rapid encystment of Acanthamoeba castellanii and survive in amoebal cysts for three weeks postinfection. Applied and environmental microbiology. 2009; 75(23):7488–500. doi: 10.1128/ AEM.01829-09 PMID: 19820161 38. Lauriano CM, Barker JR, Yoon SS, Nano FE, Arulanandam BP, Hassett DJ, et al. MglA regulates transcription of virulence factors necessary for Francisella tularensis intraamoebae and intramacrophage survival. Proceedings of the National Academy of Sciences of the United States of America. 2004; 101 (12):4246–9. PMID: 15010524 PLOS Pathogens | DOI:10.1371/journal.ppat.1005208 December 3, 2015 8/8
Computational complexity evaluation of ann algorithms for image steganalysis Computational Complexity Evaluation Of Ann Algorithms For Image Steganalysis
RECENT ACTIVITIES
Autor
Documento similar

Computational complexity evaluation of ann algorithms for image steganalysis

Livre