About this Journal Submit a Manuscript Table of Contents
BioMed Research International
Volume 2013 (2013), Article ID 578781, 7 pages
http://dx.doi.org/10.1155/2013/578781
Research Article

Chemical Composition, Nutritive Value, and Toxicological Evaluation of Bauhinia cheilantha Seeds: A Legume from Semiarid Regions Widely Used in Folk Medicine

1Departamento de Bioquímica e Biologia Molecular, Universidade Federal do Ceará, 60455-970 Fortaleza, CE, Brazil
2Departamento de Biologia, Universidade Federal do Ceará, 60455-970 Fortaleza, CE, Brazil
3Departamento de Química Orgânica e Inorgânica, Universidade Federal do Ceará, 60455-970 Fortaleza, CE, Brazil

Received 13 February 2013; Revised 4 March 2013; Accepted 11 March 2013

Academic Editor: Elvira Gonzalez De Mejia

Copyright © 2013 Daniel Câmara Teixeira et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Among the Bauhinia species, B. cheilantha stands out for its seed protein content. However, there is no record of its nutritional value, being used in a nonsustainable way in the folk medicine and for large-scale extraction of timber. The aim of this study was to investigate the food potential of B. cheilantha seeds with emphasis on its protein quality to provide support for flora conservation and use as raw material or as prototype for the development of bioproducts with high added socioeconomic value. B. cheilantha seeds show high protein content (35.9%), reasonable essential amino acids profile, low levels of antinutritional compounds, and nutritional parameters comparable to those of legumes widely used such as soybean and cowpea. The heat treatment of the seeds as well as the protein extraction process (to obtain the protein concentrate) increased the acceptance of diets by about 100% when compared to that of raw Bc diet. These wild legume seeds can be promising alternative source of food to overcome the malnutrition problem faced by low income people adding socioeconomic value to the species.

1. Introduction

According to the recent FAO database, the total number of undernourished people in the world, almost 870 million, is still unacceptably high, and eradication of hunger remains a major global challenge [1]. Hence, the search for alternative food ingredients remains of utmost importance. In this context, little-known legumes may be important in terms of food security, nutrition, agricultural development, and economic growth, standing out as important factors to reduce hunger and malnutrition in developing countries. Besides their already well-described nutritional properties, such as the abundance of macro- and micronutrients, legume seeds can display nutraceutical properties which can help to prevent heart diseases, diabetes, digestive tract diseases, and obesity [2].

As in many regions of the world, the semiarid of Brazil presents a strong contrast—on one side a high biodiversity of plant species, especially the Leguminosae family with 200 described species, and on the other, a huge population with serious public health problems that could be mitigated by the sustainable use of this biodiversity [3]. In general, the legumes are underexploited by local population, being used in a nonsustainable way in the folk medicine or for other purposes. A good example in this scenario is the species Bauhinia cheilantha (Fabales: Caesalpiniaceae). Its aerial parts have been used in the folk medicine in many communities that live in the semiarid region of Northeastern Brazil due to their hypoglycemic properties [4]. At the same time, it is already known that among Bauhinia species, B. cheilantha stands out for its seed protein content [5], but there is no record of studies on the nutritional value of these seeds. In view of this, the present work aims to investigate the food potential of B. cheilantha seeds with emphasis on its protein quality to provide support for the preservation of the regional flora and use as raw material or as prototype for the development of bioproducts with high added socioeconomic value. We thus believe that by generating basic knowledge it is possible to not only promote the preservation of the regional flora but also to use it sustainably.

2. Materials and Methods

2.1. Seed Samples and Processing

Harvesting of plant material was carried out in Caatinga forest (semiarid vegetation), Northeastern Brazil. B. cheilantha was identified and a voucher specimen deposited at Herbarium Prisco Bezerra, Universidade Federal do Ceará (UFC), under the number EAC 43701. The seeds were separated from pods and ground into fine meal (mesh size 1.0 mm) by using a blender and a coffee mill and the meal was stored at 4°C.

2.2. Proximate Composition and Amino Acid Analysis

Moisture, ash, and fat contents were assayed by the Association of the Official Analytical Chemists [6], methods 945.39A, B, and D, respectively. Nitrogen was determined using the Kjeldahl method [7] and the quantity of protein was calculated as N × 6.25. The total dietary fiber was determined by Prosky-AOAC, method 985.29 [6]. The digestible carbohydrate content was determined by difference. The energy content was estimated by multiplying the percentages of crude protein, crude fat, and digestible carbohydrates by their respective modified Atwater factors, which are 17, 37, and 17 kJ/g, respectively [8].

For amino acids analysis defatted seed meals were hydrolyzed with 6 N HCl containing 10 g/L phenol at 110°C for 22 h, in sealed glass tubes under N2 atmosphere. HCl and phenol were removed by evaporation and the amino acid compositions were established after chromatography on a Biochrom 20 system (Pharmacia LKB Biotech AB-Uppsala, Sweden). Tryptophan was measured colorimetrically by using a standard curve with concentrations ranging from 0.05 to 0.5 µg/mL of this amino acid [9].

2.3. Protein Concentrate Preparation

Seed meal was suspended (1 : 10 w/v) in 0.7 M NaCl. The suspension was maintained under continuous stirring, overnight, at 4°C, and then filtered. The filtrate was centrifuged at 12000 ×g, for 20 min, at 4°C and the clear supernatant dialyzed (cut-off 12 kDa) against distilled water. The dialyzed aqueous extract was freeze dried and called B. cheilantha protein concentrate. The total protein content was measured as previously described and total soluble proteins using the Bradford method [10].

2.4. Determination of Antinutritional and/or Toxic Protein

Haemagglutinating activity was assessed by serial twofold dilution of samples [11]. B. cheilantha crude extract was diluted with 0.15 M NaCl and mixed with rabbit erythrocytes (20 mg/mL suspension in 0.15 M NaCl), untreated or treated with trypsin (1 mg/mL). Trypsin inhibitor activity was determined using L-BAPNA as substrate [12]. Urease assay was carried out by minor modifications of the procedure described by Kaplan [13]. Briefly, for preparation of urease solution (50 µg/mL), glycerol was added to EDTA-Phosphate buffer 0.2 M, pH 6.5, at a ratio of 1 : 4 in order to preserve this solution for about three weeks, at 4°C. In addition, to obtain the standard curve, urea was used at 0.5 M instead of 0.3 M and it was prepared in the same buffer above. Acute toxicity to mice was verified by intraperitoneal injection of samples [14]. All animal experiments were approved by the Animal Experimentation Ethics Committee of the Universidade Federal do Ceará (CEPA/UFC).

2.5. Susceptibility of Antinutritional and/or Toxic Proteins to In Vitro Digestion

B. cheilantha protein concentrate (20 mg) was dissolved in 1 mL of pepsin (0.1 mg/mL in 0.1 N HCl, pH 1.8). The mixture was incubated at 37°C, for 2 h. After that, 0.5 mL was removed and an equal volume of 0.25 M Tris-HCl, pH 8.9, added to inactivate pepsin action and then submitted to haemagglutinating, urease, and trypsin inhibitory activities. The remaining 0.5 mL was added to 0.5 mL of trypsin and chymotrypsin (0.1 mg/mL of each enzyme in 0.25 M Tris-HCl, pH 8.9) and kept at 37°C, for 3 h. The mixture was tested for the same activities. The enzyme solutions were prepared to provide an enzyme/sample relation of 1 : 100 (w/w).

2.6. Diets

Four diets containing B. cheilantha were prepared: (1) Raw B. cheilantha diet (Raw Bc); (2) Soaked B. cheilantha diet (Soaked Bc); (3) Soaked and heat treated B. cheilantha diet (Heated Bc); and (4) B. cheilantha protein concentrate diet (BcPCd). To obtain soaked Bc diet, seeds were soaked for 1 h in distilled water (1 : 4, w/v) prior to grinding. The heat treatment consisted of boiling at 92°C for 60 min. This was sufficient to abolish more than 50% of urease and trypsin inhibitory activities and 100% of haemagglutinating activity. After boiling, the seeds along with the cooking water were placed in an oven at 37°C, for 24 h, followed by subsequent grinding. For comparison purposes, casein and soybean diets were offered. Diets containing B. cheilantha were supplemented with L-methionine based on the amino acid contents of seeds and protein concentrate, to meet the amino acid requirements of rats [15]. A nonprotein containing diet (NPC) was fed to allow determination of some nutritional parameters (Table 1).

tab1
Table 1: Composition (g/Kg) of casein, soybean, NPC, and experimental diets.
2.7. Feeding Trials

Male Wistar rats were weaned at 21 days of age and given a commercial stock diet until their weights reached around 50 g. They were then fed the casein diet ad libitum for 3 days as a period of adaptation. Groups of 6 rats were placed in individual cages where they received water and diets ad libitum for 10 days. Rat weights, diet spillage, and refused diet were recorded daily. During the last 5 days of the experiment, feces were collected, freeze dried, weighed, and ground in a coffee grinder. At the end of the trial, the rats were killed by halothane overdose. The carcasses were dried in oven at 100°C for 72 h and ground. The nitrogen content of carcasses and feces was measured as described previously. True protein digestibility and net protein utilization (NPU) were calculated as described by Vasconcelos et al. [14].

2.8. Phytochemical Compounds

The diets were screened for phytochemical compounds such as phenols and tannins (reaction with ferric chloride), leucoanthocyanidins (heat treatment followed by alkalinization and acidification of sample), flavonoids and xanthones (reaction of magnesium granules with hydrochloric acid), steroids and triterpenes (extraction with chloroform, acetic anhydride, and sulfuric acidic), saponins (foam production after water solubilization and stirring), and alkaloids (precipitation with Hagger, Mayer and Dragendorff reagents) according to Matos [16]. These analyses were based on visual observation of color modification or precipitate formation after addition of specific reagents.

2.9. Statistical Analyses

Data were subjected to one-way analysis of variance (ANOVA) and the significance of differences among means was determined using the Tukey test ( ).

3. Results and Discussion

Regarding the proximate composition (Table 2), B. cheilantha seeds (raw, soaked, or heated) showed high protein content, which exceeds the levels found in other seeds of the same genus (16.8 to 29.3%) [17]. The protein content of B. cheilantha seeds is comparable or much higher than those reported for some important cultivated legume seeds such as soybean (39.5 to 44.5%) and cowpea (19.5 to 26.1%) [18, 19]. Often poor populations in the developing world depend on legume seeds, particularly beans for protein intake, mainly because the animal protein, that is considered more complete than that of plant source, is usually more expensive. In this context, the search for unconventional protein sources is very timely to analyze the feasibility of their incorporation into the diet as a way to combat the nutritional deficiencies of the poorest populations, resulting from the intake of foods that do not reach nutritional requirements [3].

tab2
Table 2: Proximate composition of B. cheilantha seed meal expressed as a percentage of dry matter.

The seeds showed also high content of dietary fiber, which was higher than those described for legume seeds widely consumed, such as soybean (17.4%) and black beans (22.6%) [20]. This is an interesting finding since the consumption of dietary fiber has been related to prevention of cardiovascular disease, diabetes, and digestive tract diseases, considering that it lowers the glycemix index of food as well as serum cholesterol levels [2, 21]. The detected levels of moisture, ash, lipids, and digestible carbohydrates were lower or similar to those described for legumes [3].

The levels of essential amino acids (Table 3) in the whole meal of B. chemilantha were lower than that of casein [22] and did not reach the requirement for rats [15]. Similar values were observed for the processed seeds. This fact was also demonstrated for soybean seeds [14]. Nevertheless, the essential amino acid levels of B. cheilantha protein concentrate were superior for valine, methionine + cysteine, and tryptophan and met the recommendations for 2–5 and 10–12 years old children [23], when compared to protein isolates from cowpea [24], which is one important plant protein source in Latin America as well as in many other regions of the world. This is of utmost importance considering that usually legume seeds are low in sulfur-containing amino acids (<2%) and tryptophan (<1%) [25].

tab3
Table 3: Amino acid composition (g/100 g protein) of B. cheilantha meal seed and protein concentrate compared to amino acid requirements for children and growing rats.

Antinutritional/toxic activities are depicted in Table 4. The crude extract of B. cheilantha seeds presented trypsin inhibitory activity (31.5 gTI/Kg) similar to those of some soybean (28.5 to 62.5 gTI/Kg) and cowpea (12.0 to 30.6 gTI/Kg) genotypes [14, 19]. This activity was almost completely abolished after 90 min boiling. The haemagglutinating activity (793.6 HU/gF) determined in the seed extract was the same for both untreated and trypsin treated rabbit erythrocytes and lower than those found for soybeans (1152 to 18432 HU/gF) [18]. The heat treatment of seeds for 30 min was able to inactivate the ability to agglutinate red cells. The urease activity for the raw seeds (32.9 U/gF) is lower than that reported for soybean seeds (107.3 to 219.3 U/gF) [14]. As to acute toxicity of B. cheilantha extract, it was not lethal when injected intraperitoneally in mice. The low content of toxic and/or antinutritional factors adds positive attributes favoring the use of B. cheilantha seeds, considering that many important plant sources of protein and fiber (beans and soybeans) contain these compounds in higher quantities.

tab4
Table 4: Trypsin inhibitory, lectin, urease, and toxic activities present in the crude extract from B. cheilantha seeds and their stability to heat treatment.

Regarding the susceptibility of the antinutritional and/or toxic proteins to in vitro digestion, the haemagglutinating and trypsin inhibitory activities were not detected after digestion of seed extract with pepsin. On the other hand, the urease activity was reduced from 32.9 to 10.3 U/gF. However, when the seed extract was subjected to sequential digestion with pepsin, trypsin, and chymotrypsin, there was no detected urease activity indicating that this protein only may exert its effect in the gastric mucosa of rats (data not shown).

Considering the nutritional parameters (Table 5), the growth rate of the groups fed on raw Bc and soaked Bc was significantly similar to that observed for the NPC group. However, the animals fed on BcPC and heated Bc showed growth rate higher than those of raw Bc diet. The body weights of rats fed on heated Bc were significantly similar to that of soybean group but lower than those fed on casein diet. The dietary intake in the experimental groups was equivalent to about 25 to 50% of that in casein group, which must have impaired animals growth. Feeding studies have showed that raw legumes did not support the growth of rats [26]. It is well known that rats reduce considerably their intake when the diet is poor in protein or has low quality proteins, which does not seem to be the case in this study. Another factor which could have interfered with food intake is the high dietary fiber content of the B. cheilantha seeds, which is well above 30%. In fact, it is known that dietary fiber may affect gastric emptying since it may slow gastric filling, due to its bulking and energetic dilution capacity [27]. The slow gastric emptying in turn reduces food intake. However, this does not seem to occur since dietary fiber contents of raw Bc and heated Bc were similar to each other whereas the food intake of rats on heated Bc was two-fold higher than those of raw Bc. Besides, the food intake of rats on heated Bc was similar to those of BcPC which does not contain fiber. Similarly, the organoleptic properties of the diets can cause significant impairment of dietary intake [28]. It is likely that the poor organoleptic properties of the diets based on B. cheilantha meal and protein concentrate are responsible for the low dietary intake.

tab5
Table 5: Nutritional parameters of rats fed on B. cheilantha seed meal and protein concentrate compared with those of rats fed on NPC, casein, and soybean diets.

The NPU values (Table 5) for groups fed on raw Bc and soaked Bc diets were negative. However, the NPU values of the heated Bc and BcPC diets were significantly similar to that of the diet based on soybean. The diet consisting of BcPC showed protein digestibility value significantly similar to those of casein and soybean. The digestibility values of raw, soaked, and heated Bc diets were higher than those for raw cowpea (46.5–60.3%) [29]. Raw and soaked Bc diets had negative biological values whereas heated Bc (33.5%) and BcPC diets (29.2%) did not differ significantly from that of soybean diet (28.5%). Concerning the parameters of protein quality (NPU, digestibility, and biological value), the results of the groups fed on heated Bc and BcPC diets were promising, contrary to those of the treatments with raw Bc and soaked Bc diets. In general, the heat treatment of the seeds as well as the protein extraction process (to obtain the protein concentrate) increased the acceptance of diets by about 100% when compared to that of raw Bc diet.

As an attempt to comprehend the poor acceptance of raw Bc diets, phytochemical compounds were analyzed and tannins, flavonoids, xanthones, triterpenoids, saponins, and steroids were detected. In the other test diets only saponins and steroids were observed. All these compounds have been associated with poor palatability [28] and the removal of some of them must have improved the acceptance of the heated Bc and BcPC diets. However, further studies are necessary to clarify which of these compounds must be responsible for the poor acceptance of the raw Bc diet by the animals.

4. Conclusion

These wild legume seeds can be promising alternative source of food to overcome the malnutrition problem faced by low income people, as well as to create basic sustainability elements to prevent extinction of this species. B. cheilantha seeds show high protein content, reasonable essential amino acids profile, low levels of antinutritional compounds, and protein quality parameters comparable to those of legumes widely used such as soybean and cowpea. Nevertheless, their organoleptic properties should be improved by technological processes, such as heating and developing protein concentrates, in order to use the full nutritional potential of these seeds.

Acknowledgment

This work was supported by CNPq, CAPES, and FUNCAP.

References

  1. FAO, The State of Food Insecurity in the World, Food and Agriculture Organization of the United Nations, Rome, Italy, 2012.
  2. V. Vadivel, A. Nandety A, and H. K. Biesalski, “Antioxidant potential and health relevant functionality of traditionally processed Cassia hirsuta L. seeds: an Indian underutilized food legume,” Plants Foods for Human Nutrition, vol. 66, pp. 245–253, 2011.
  3. A. F. U. Carvalho, D. F. Farias, L. C. B. da Rocha-Bezerra et al., “Preliminary assessment of the nutritional composition of underexploited wild legumes from semi-arid Caatinga and moist forest environments of northeastern Brazil,” Journal of Food Composition and Analysis, vol. 24, no. 4-5, pp. 487–493, 2011. View at Publisher · View at Google Scholar · View at Scopus
  4. S. L. Cartaxo, M. M. de Almeida Souza, and U. P. de Albuquerque, “Medicinal plants with bioprospecting potential used in semi-arid northeastern Brazil,” Journal of Ethnopharmacology, vol. 131, no. 2, pp. 326–342, 2010. View at Publisher · View at Google Scholar · View at Scopus
  5. R. Braga, Plantas do Nordeste Especialmente do Ceará, Imprensa Oficial, Fortaleza, Brazil, 2nd edition, 1960.
  6. AOAC, Official Methods of Analysis, Association of Official Analytical Chemists, Washington, DC, USA, 16th edition, 1997.
  7. W. E. Baethgen and M. M. Alley, “A manual colorimetric procedure for measuring ammonium nitrogen in soil and plant Kjeldahl digests,” Communications in Soil Science and Plant Analysis, vol. 20, pp. 961–969, 1989.
  8. FAO, “Food Energy: Methods of Analysis and Conversion Factors,” Report of a Technical Workshop, FAO, Rome, Italy, 2003.
  9. M. Pintér-Szakács and I. Molnár-Perl, “Determination of tryptophan in unhydrolyzed food and feedstuffs by the acid ninhydrin method,” Journal of Agricultural and Food Chemistry, vol. 38, no. 3, pp. 720–726, 1990. View at Scopus
  10. M. M. Bradford, “A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding,” Analytical Biochemistry, vol. 72, no. 1-2, pp. 248–254, 1976. View at Scopus
  11. R. A. Moreira and J. C. Perrone, “Purification and partial characterization of a lectin from Phaseolus vulgaris,” Plant Physiology, vol. 59, no. 5, pp. 783–787, 1977.
  12. G. H. Hamerstrand, L. T. Black, and J. D. Glover, “Trypsin inhibitors in soy products: modification of the standard analytical procedure,” Cereal Chemistry, vol. 58, pp. 42–45, 1981.
  13. A. Kaplan, “The determination of urea, ammonia, and urease,” Methods of Biochemical Analysis, vol. 17, pp. 311–324, 1969. View at Scopus
  14. I. M. Vasconcelos, A. A. B. Maia, E. A. Siebra et al., “Nutritional study of two Brazilian soybean (Glycine max) cultivars differing in the contents of antinutritional and toxic proteins,” Journal of Nutritional Biochemistry, vol. 12, no. 1, pp. 55–62, 2001. View at Publisher · View at Google Scholar · View at Scopus
  15. M. E. Coates, P. N. Odonoghue, P. R. Payne, and R. J. Ward, Dietary Standards for Laboratory Rats and Mice: Nutritional and Microbiological Recommendation, Laboratory Animal Handbook 2, London, UK, 1969.
  16. F. J. A. Matos, Introdução a Fitoquímica Experimental, EUFC, Fortaleza, Brazil, 2nd edition, 1999.
  17. N. Rajaram and K. Janardhanan, “Chemical composition and nutritional potential of the tribal pulses Bauhinia purpurea, B. racemosa and B. vahlii,” Journal of the Science of Food and Agriculture, vol. 55, pp. 423–431, 1991.
  18. I. M. Vasconcelos, C. C. Campello, J. T. A. Oliveira, A. F. U. Carvalho, D. O. B. Sousa, and F. M. M. Maia, “Brazilian soybean Glycine max (L.) Merr. cultivars adapted to low latitude regions: seed composition and content of bioactive proteins,” Revista Brasileira de Botanica, vol. 29, no. 4, pp. 617–625, 2006. View at Publisher · View at Google Scholar · View at Scopus
  19. F. M. M. Maia, J. T. A. Oliveira, M. R. T. Matos, R. A. Moreira, and I. M. Vasconcelos, “Proximate composition, amino acid content and haemagglutinating and trypsin-inhibiting activities of some Brazilian Vigna unguiculata (L) Walp cultivars,” Journal of the Science of Food and Agriculture, vol. 80, no. 4, pp. 453–458, 2000. View at Scopus
  20. L. C. Trugo, C. M. Donangelo, N. M. F. Trugo, and K. E. Bach Knudsen, “Effect of heat treatment on nutritional quality of germinated legume seeds,” Journal of Agricultural and Food Chemistry, vol. 48, no. 6, pp. 2082–2086, 2000. View at Publisher · View at Google Scholar · View at Scopus
  21. B. Ruiz-Roso, J. C. Quintela, E. de la Fuente, J. Haya, and L. Pérez-Olleros, “Insoluble carob fiber rich in polyphenols lowers total and ldl cholesterol in hypercholesterolemic sujects,” Plant Foods for Human Nutrition, vol. 65, no. 1, pp. 50–56, 2010. View at Publisher · View at Google Scholar · View at Scopus
  22. P. G. Reeves, F. H. Nielsen, and G. C. Fahey, “AIN-93 purified diets for laboratory rodents: final report of the American Institute of Nutrition ad hoc writing committee on the reformulation of the AIN-76A rodent diet,” Journal of Nutrition, vol. 123, no. 11, pp. 1939–1951, 1993. View at Scopus
  23. FAO/WHO/UNU, “Energy and Protein Requirements,” Technical Report Series 724, World Health Organization, Geneva, Switzerland, 1985.
  24. A. Rangel, K. Saraiva, P. Schwengber et al., “Biological evaluation of a protein isolate from cowpea (Vigna unguiculata) seeds,” Food Chemistry, vol. 87, no. 4, pp. 491–499, 2004. View at Publisher · View at Google Scholar · View at Scopus
  25. K. Gallardo, R. Thompson, and J. Burstin, “Reserve accumulation in legume seeds,” Comptes Rendus—Biologies, vol. 331, no. 10, pp. 755–762, 2008. View at Publisher · View at Google Scholar · View at Scopus
  26. L. A. Rubio, G. Grant, S. Bardocz, P. Dewey, and A. Pusztai, “Nutritional response of growing rats to faba beans (Vicia faba L.,minor) and faba bean fractions,” British Journal of Nutrition, vol. 66, no. 3, pp. 533–542, 1991. View at Scopus
  27. N. W. Read and M. A. Eastwood, “Gastro-intestinal physiology and function,” in Dietary Fibre. A Component of Food, T. F. Schweizer and C. A. Edwards, Eds., pp. 103–117, Springer, London, UK, 1992.
  28. C. Martínez-Villaluenga, G. Urbano, J. M. Porres, J. Frias, and C. Vidal-Valverde, “Improvement in food intake and nutritive utilization of protein from Lupinus albus var. multolupa protein isolates supplemented with ascorbic acid,” Food Chemistry, vol. 103, no. 3, pp. 944–951, 2007. View at Publisher · View at Google Scholar · View at Scopus
  29. S. Y. Giami, “Compositional and nutritional properties of selected newly developed lines of cowpea (Vigna unguiculata L. Walp),” Journal of Food Composition and Analysis, vol. 18, no. 7, pp. 665–673, 2005. View at Publisher · View at Google Scholar · View at Scopus