Journal of Medical Engineering & Technology, Vol. 30, No.

6, November/December 2006, 390 – 396

Self-localization of laser induced tumour coagulation limited by heat diffusion through active tissue
B. Y. DATSKO*{, V. V. GAFIYCHUK{, I. A. LUBASHEVSKYx and A. V. PRIEZZHEV{ {Institute for Applied Problems in Mechanics and Mathematics, National Academy of Sciences of Ukraine, Naukova Street 3b, 79053, Lviv, Ukraine {Institute of Computer Modeling, Krakow University of Technology, 24 Warszawska Street, 31155, Krakow, Poland xTheory Department, General Physics Institute, Russian Academy of Sciences, 38, Vavilov Street, Moscow, Russia {Physics Department, Moscow State University, Vorobjovy Gory, 119899, Moscow, Russia
We analyse necrosis growth due to thermal coagulation induced by laser light absorption and limited by heat diffusion into the surrounding live tissue. The tissue is assumed to contain a tumour in the undamaged tissue where the blood perfusion rate does not change during the action. By contrast, normal tissue responds strongly to an increase in the tissue temperature and the blood perfusion rate can grow by tenfold. We study in detail necrosis formation under conditions typical of a real course of thermal therapy treatment. The duration of the treatment is about 5 minutes when a necrosis domain of about 1 cm or above is formed. In particular, if the tumour size is sufficiently large, i.e. it exceeds 1 cm, and the tissue response is not too delayed, i.e. the delay time does not exceed 1 min, then there are conditions under which the relative volume of the damaged normal tissue is small in comparison with the tumour volume after the tumour is totally coagulated. Keywords: Necrosis growth; Thermal coagulation; Tumour; Perfusion

1. Laser induced heat diffusion limited tissue coagulation Thermal coagulation of living tissue caused by local heating due to laser light absorption is one of the novel thermotherapy techniques of tumour treatment which is currently undergoing clinical trials (see e.g. [1]). Thermal coagulation is used to form a necrosis domain with a desired form for the removal of the malignant tissue. So mathematical modelling of the necrosis growth is required, first, to find out the physical limitations and the basic characteristics of the treatment and, second, to optimize the therapy course. However, living tissue is extremely complex in structure, so for an adequate theoretical model to be developed and for the mathematical modelling of the given process to be implemented

reliably, typical limit cases should be singled out and studied individually. One of these cases is laser induced heat diffusion limited tissue coagulation [2,3]. It is characterized by the following features (figure 1). Absorption of laser light delivered into a small internal region of living tissue causes the temperature to attain high values (about or above 708C), leading to immediate thermal coagulation in this region. Heat diffusion into the surrounding live tissue causes further coagulation, giving rise to the growth of the necrosis domain. In this case heat diffusion plays a significant role in necrosis growth because the necrosis size < exceeds the depth of laser light penetration into the tissue. Therefore, the temperature distribution inevitably has to be substantially nonuniform and for the tissue to coagulate at peripheral points heat

*Corresponding author. Email: vagaf@yahoo.com
Journal of Medical Engineering & Technology ISSN 0309-1902 print/ISSN 1464-522X online Ó 2006 Informa UK Ltd. http://www.tandf.co.uk/journals DOI: 10.1080/03091900500467340

Self-localization of coagulation in active tissue

391

Figure 1. Necrosis growth due to local thermal tissue coagulation limited by heat diffusion in the surrounding undamaged tissue. d ¼ 1, 2 corresponds to the surface plane applicator and the internal cylindrical one. diffusion should cause the temperature to grow at these points. The latter property singles out the specific mode of thermal coagulation under discussion from other possible types of thermotherapy treatment, and that is why we refer to necrosis growth under the given conditions as thermal tissue coagulation limited by heat diffusion. In particular, it turns out that optimal implementation of this thermotherapy mode is characterized by the formation of necrosis domains of size < 0 1 cm and by a treatment duration of tcourse*<2/D*2 – 8 min (where D % 2610 – 3 cm2 s – 1 is the tissue temperature diffusivity) [2,3]. Heat diffusion in the live tissue is affected substantially by blood perfusion causing the heat sink [4]. Thus, the characteristics of the spatial distribution and the dynamics of the blood perfusion rate should have a substantial effect on the necrosis growth limited by heat diffusion. Therefore, in modelling this therapy mode one has to take into account the tissue response to the temperature growth, which can locally give rise to a tenfold increase in the blood perfusion rate [5]. The latter effect, in particular, is responsible for a substantially nonuniform distribution of the blood perfusion rate and visually manifests itself in a red ring (‘hyperemic ring’) appearing around the necrosis region during the thermotherapy treatment. In previous papers based on the free boundary description [6 – 13], we developed a model for heat diffusion limited thermal tissue coagulation, and studied the properties of the corresponding necrosis growth. In particular, we have shown that the given mode of thermal coagulation comprises two stages, fast and slow. At the former the necrosis domain grows fast and its size attains values of order < * 1 cm. The latter is characterized by a substantially slower growth of the necrosis domain. At this stage the necrosis growth is governed by the competition between the heat diffusion into the surrounding undamaged tissue

and the heat dissipation caused by blood perfusion. Therefore, a sufficiently strong tissue response to temperature variations which, in turn, gives rise to an essential local increase in the blood perfusion rate, has to affect the necrosis formation. Therefore, the higher the local blood perfusion rate, the smaller is the size of the necrosis domain. Inside tumours the blood vessels are abnormal; in particular, they have lost the ability to expand substantially in the response to tissue heating. As a result, inside a tumour the blood perfusion rate can increase no more then twice under strong local heating, although in the unaffected tissue the blood perfusion rate inside a tumour typically exceeds the perfusion rate in the normal tissue [5]. So, if normal tissue exhibits a strong response to temperature variations, then the boundary of the necrosis domain will penetrate deeper inside the tumour than in the normal tissue during the thermal therapy course. Therefore in this case, first, the relative volume of the damaged normal tissue can be less than the volume of the tumour after its coagulation. Second, when the necrosis boundary reaches the tumour boundary the necrosis growth should slow down. At first glance it would seem that the necrosis growth is mainly confined to the tumour space. This phenomenon is the subject of the present paper and will be called the selflocalization of the necrosis growth in active living tissue with a tumour. 2. Free boundary model of the necrosis growth We study necrosis formation caused by laser light absorption in living tissue with a tumour, by applying the following model [2,6 – 13]. Laser light absorption causes Heat generation and, as a result, thermal coagulation of the tissue. For simplicity’s sake the heat generation rate q(r) is considered to be a given function of the spatial coordinate r that is independent of the tissue state and the time t (certainly, for t 4 0, where t ¼ 0 is the beginning of the action on the tissue). In addition, we assume that the heat generation rate q(r) is localized in a small region adjacent to the applicator, whose thickness is substantially less than the characteristic size < 0 1 cm of the necrosis domain formed during a typical course of thermotherapy treatment (figure 2). In modelling heat propagation in the tissue we single out three regions: the necrosis domain Qn, where the blood perfusion rate is equal to zero: jðr; tÞ ¼ 0 for r 2 Qn ; ð1Þ

the undamaged tumour tissue Qt, and the undamaged (live) normal tissue Ql. Inside the necrosis domain Qn the tissue temperature T(r, t) obeys the heat diffusion equation for solids: cr @T ¼ kr2 T þ q; @t r 2 Qn ; ð2Þ

392

B. Y. Datsko et al.

Figure 2. The system where the necrosis formation has been studied numerically based on the free boundary model. The xyregion of the plane and cylindrical symmetry imitates the effect of the surface applicator and that of the internal cylindrical one. pffiffiffiffiffiffi where ln $ 1= Ln is also a constant of order unity. Equation (6) should be subjected to a certain boundary condition at the interface À of the necrosis domain because it makes no sense to average the blood perfusion rate over the necrosis domain impermeable to blood. The physical layer separating the necrosis domain and the undamaged tissue where the local vascular network is not damaged is complex in structure and contains a spatial increase of the blood perfusion rate from zero to the value in the undamaged tissue. In order to avoid the problem of analysing the blood perfusion rate in this layer we take into account the following simplifying circumstance. On one hand, the typical size of the necrosis domain formed during a thermal therapy course and the characteristic length of temperature variations are of the same order: about 1 cm. So, particular details of spatial variations in the blood perfusion rate on scale much less than 1 cm are not a factor. On the other hand, the damaged part of the vascular network located inside the necrosis domain is most likely to be made up of an artery and vein having previously supplied this region with blood as a whole, and of shorter vessels formed by their branching. Therefore, the region containing the vascular network part in which blood flow is remarkably disturbed because of the necrosis formation does not exceed substantially the necrosis domain. The latter enables us not to differentiate between the given layer and the interface À and to deal with a sharp jump of the blood perfusion rate at the necrosis interface. The desired boundary condition imposed on the averaged blood perfusion rate jn meets the requirement that the normal gradient of the averaged blood perfusion rate be equal to zero at the interface À: rn jn j À ¼ 0: ð7Þ

where c and r are the specific heat and density of the tissue, and k is the cellular tissue heat conductivity. In the undamaged tissue (including also the undamaged part of the tumour) the temperature is governed by the bioheat equation (see also [14]): cr @T ¼ Fkr2 T À fcrjn ðT À Ta Þ þ q; @t r 2 Ql [ Qt : ð3Þ

Here Ta is the temperature of arterial blood in systemic circulation, jn is the value of the blood perfusion rate j locally averaged over spatial scales ‘n ðrÞ $ rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi k $ 4 mm; crf jv ðrÞLn ð4Þ

where Ln ¼ ln(l/a) % 3 – 4, l/a is the mean ratio of the individual length to radius of blood vessels forming pffiffiffiffiffiffi peripheral circulation, the cofactor f $ 1= Ln $ 0:5 takes into account the counter-current effect [15,16], and the factor F ¼ F(Ln)*2 allows for the renormalization of the heat conductivity due to blood flow. It should be pointed out that the scale ‘n(r) is the minimal scale on which equation (3) is justified [14] and the characteristic size of the necrosis domain formed during this mode of thermal pffiffiffiffiffiffi therapy can be estimated as < 0 Ln ‘n : rffiffiffiffiffiffiffiffiffi k : ð5Þ <0 crf j Finding the relationship between the averaged and true blood perfusion rates, jn and j, we have to take into account that the scale ‘n of averaging in its turn depends on the local value of jn (expression (4)). This dependence enables us to specify this relationship in the form [14]: jn À ln k 2 r ln jn ¼ j; cr r 2 Ql [ Qt ; ð6Þ

We note that the adopted boundary condition will not hold if a large vessel goes through the necrosis domain.

Self-localization of coagulation in active tissue

393

However, the probability of this event is small and this case should be analysed individually. Now let us specify the tissue response to temperature variations. Only the undamaged normal tissue is considered to respond to local heating, inside the undamaged tissue of the tumour the blood perfusion rate maintains the initial value j0 assumed for simplicity’s sake to be the same in the normal tissue and the tumour: jðr; tÞ ¼ j0 for r 2 Qt : ð8Þ

pig liver at T0 ¼ 608C). Below in numerical calculations the dependence (13) will be taken in the form:   T À 60 minÀ1 ; ð14Þ oðTÞ ¼ 0:2 exp 3:6 where the temperature T is in 8C.

3. Results of numerical simulation: self-localization of the necrosis growth in living tissue with a tumour The stated model has been used to analyse numerically the necrosis growth in the case shown in figure 2. The laser light absorption near the applicator gives rise to the heat generation rate q(y) uniform along the x-axis and decreasing with y as q(y) ! exp(7y/lad), where the characteristic thickness lad of the laser light absorption layer is chosen less then the characteristic size < 0 1 cm of the necrosis domain formed during the thermal therapy mode under consideration, lad 5 <. The intensity of laser light has taken such values that the tissue temperature directly near the applicator, y ¼ 0, be about 80 – 908C (q(y) ¼ 2 exp(7y) for large tumour and q(y) ¼ 4 exp(74y) for small tumour). We have studied two opposite cases specified by the relation of the scale < (see expression (5)) and the tumour size L. For a small tumour, L 5 <, the self-localization phenomenon cannot come into being and this case is presented for illustration only. For a large tumour, L 4 <, there are conditions under which the self-localization is pronounced, as is demonstrated directly in the present section. Besides, we have considered the xy-region of plane geometry as well as the cylindrical one in order to compare necrosis growth for applicators of different geometry—surface plane applicator and internal cylindrical. In the latter case the radius r0 of the applicator was also assumed to be small in comparison with the scale <. In numerical simulation we have used the following typical values of the thermal conductivity: k*761073 W cm71 K71, the heat capacity c*3.5 J g71 K71, and the density r*1 cm73 of the tissue, as well as set the blood perfusion rate j0*0.3 min71 and the factors f ¼ 0.5, F ¼ 2. Figure 3 demonstrates the necrosis growth in living tissue without response. In this case the existence of the tumour has no effect on the necrosis growth. A sufficiently small tumour also cannot substantially affect the necrosis growth, as is seen in figure 4, which shows necrosis growth in living tissue with strong response without delay. The tumour radius was chosen to be equal to L ¼ 0.5 cm, and the tissue response to heating was assumed to be strong, a ¼ 0.1, and without delay, t ¼ 0. This case has been considered because under such conditions the effect of a large tumour on necrosis growth is most pronounced. When the tumour size is sufficiently large, L 4 <, and the tissue response is not too delayed, t ( <2 /D the self-localization phenomenon comes into being as

Blood vessels can expand only to a certain extent as the temperature grows. So when it becomes sufficiently high, T 4 Tvr, the blood perfusion rate j(r, t) attains a large but finite value, jmax, and remains approximately constant. Keeping in mind this feature we describe the normal tissue response to local heating by the equation: t @j þ jFðTÞ ¼ j0 @t for r 2 Ql : ð9Þ

Here t is the delay time of the tissue response and the function F(T) is of the form: ( FðTÞ ¼ a þ ð1 À aÞ Tvr À T Tvr À Ta a for T < Tvr for T > Tvr

;

ð10Þ

where a ¼ j0/jmax and the temperature at which the blood vessels exhaust their ability to expand is estimated as Tvr % 45 – 468C. Equations (2) and (3) governing the evolution of the tissue temperature should be completed by the boundary conditions relation the temperature field and the necrosis interface À. To solve this problem we have developed a free boundary model that assumes the temperature and heat flux to have no jumps, i.e. the temperature distribution meets the following boundary conditions: T j Àþ0 ¼ T j ÀÀ0 ¼ Tcg ;
def

F rn T j Àþ0 ¼ rn T j ÀÀ0 ;

ð11Þ

and the normal velocity of the interface À is given by the expression:   =0 o0 D Tcg À T0 : ð12Þ exp #n ¼ jrn TjÀÀ0 D Here the constant I0*1 and the function   T À T0 oðTÞ ¼ o0 exp D

ð13Þ

are actually a convenient approximation of the Arrhenius dependence of the thermal coagulation rate oðTÞ ! exp fÀ Eg. The available experimental data [17] T for the temperature dependence of the exposure time enable us to estimate the value of D as D*3 – 58C (D ’ 3.68C for

394

B. Y. Datsko et al.

Figure 3. Necrosis growth in living tissue without response, where the existence of the tumour has no effect on the necrosis formation. The dotted lines show the position of the necrosis interface at different moments of time.

Figure 4. Necrosis growth in living tissue containing a small tumour. The tissue exhibits a strong response without delay (a ¼ 0.1, t ¼ 0). The dotted lines show the position of the necrosis interface at different moments of time.

demonstrated in figures 5 – 6. Figure 5 shows necrosis growth in living tissue containing a tumour of size of L ¼ 1.5 cm, where the tissue response is strong, a ¼ 0.1, and not delayed, t ¼ 0. This situation exactly matches the conditions under which self-localization is most pronounced. When the tissue response is delayed the effect of the tumour existence is weakened, as show in figure 6, where we set t ¼ 1 min. 4. Discussion and closing remarks Considering the treatment conditions, we would first of all like to discuss the size of the tumours that our model is applicable to. For mathematical modelling we used a tumour with a diameter of 1 – 38 cm. At these scales selflocalization of the laser induced coagulation can take place, and as a result, a lower powered heating source is needed for coagulation of the malignant tissue. The Nd:YAG laser

is commonly applied through optical fibres to treat internal organs. This laser, emitting light at 1064 nm, has been the most widely used laser for thermotherapy. It is employed in many surgical disciplines for ablation and cutting. Let us evaluate the possible parameters of the process. The wavelength of the laser light determines the level of penetration of the light into the tissue, and the absorbtion coefficient b ¼ 1 cm71 [18]. We do not consider the microscopic process of absorbing light by chromophores in living tissue. We only consider the macroscopic absorption coefficient for electromagnetic waves, taking into account the fact that near-infrared light exhibits the greatest penetration depth obtainable with laser light. In this case, a deeper penetration of the light enables a higher power deposition, and thus a greater volume of necrosis. The temperature reached at y ¼ 0 has to approximately equal 728C (during computer simulation we used q ¼ 2 at y ¼ 0). In this case we can calculate irradiance j ¼ qcrb71 T, which is equal to *200 W cm2. With the use of the corresponding fibre optics diameter emitting light from the area *0.1 cm2, the laser effect can be reached through the average irradiation of 20 W, which corresponds to the Nd:YAG laser. In this case, we do not distinguish between continuous or pulsed power irradiation because the temperature field averages on the scale of several microseconds, and only the total absorbed energy determines the level of heating of the tissue. Analysing the experimental situation, we can state that similar conditions could be realized for the treatment of breast carcinoma presented in [19]. Analysing this clinical investigation we can conclude that the case described in [19] is quite similar to the conditions we consider in this paper. It should be noted that we did not consider that the tissue can be cooled from outside because this case should be viewed from a threedimensional perspective [20,21]. In conclusion, we have analysed the necrosis growth due to thermal coagulation induced by the laser light absorption and limited by heat diffusion into the surrounding

Self-localization of coagulation in active tissue

395

Figure 5. Necrosis growth in living tissue containing a large tumour. The tissue response to temperature increase is strong and not delayed (a ¼ 0.1, t ¼ 0). The dotted lines show the position of the necrosis interface at different moments of time.

Figure 6. Necrosis growth in living tissue containing a large tumour. The tissue response to temperature increase is strong and delayed (a ¼ 0.1, t ¼ 1 min). The dotted lines show the position of the necrosis interface at different moments of time.

living tissue. The tissue is assumed to contain a certain tumour, and it is assumed that the blood perfusion rate does not change during the action. The latter assumption certainly concerns only the undamaged part of the tumour. By contrast, the normal tissue strongly responds to the increase of the tissue temperature and, as a result, the blood perfusion rate can grow tenfold. We studied in detail the necrosis formation under conditions typical of a real course of thermal therapy treatment. The duration of the action has been taken to be about 5 minutes when a necrosis domain of size about or above 1 cm is formed. In particular, we have shown that if the tumour size is sufficiently large, i.e. it exceeds 1 cm, and the tissue response is not too delayed, i.e. it does not exceed 1 min, then there are conditions under which the relative volume of the damaged normal tissue is small in comparison with the tumour volume after the tumour is totally coagulated. References
[1] Muller, G. and Roggan, A., (Eds), 1995, Laser-Induced Interstitial ¨ Thermotherapy (Bellingham, Washington: SPIE Optical Engineering Press).

[2] Lubashevsky, I.A., Priezzhev, A.V. and Gafiychuk, V.V., 2000, Laser induced heat diffusion limited tissue coagulation as a laser therapy mode. In: D.D. Duncan, J.O. Hollinger and S.L. Jacques (Eds) Laser – Tissue Interaction XI: Photochemical, Photothermal and Photomechanical, Proceedings of SPIE, 3914, 66 – 74. [3] Lubashevsky, I.A., Gafiychuk, V.V. and Priezzhev, A.V., 2001, Laser induced heat diffusion limited tissue coagulation: problem and general properties. Available online at: http://xxx.lanl.gov/abs/physics/0101002. (Accessed 01.11.1999). [4] Chen, M.M. and Holmes, K.R., 1980, Microvascular contributions in tissue heat transfer. Annals of the New York Academy of Sciences, 335, 137 – 154. [5] Song, C.W., Lokshina, A., Rhee, I.G., Patten, M. and Levitt, S.H., 1984, Implication of blood flow in hyperthermia treatment of tumours. IEEE Transactions on Biomedical Engineering, BME-31, 9 – 15. [6] Lubashevsky, I.A., Priezzhev, A.V., Gafiychuk, V.V. and Cadjan, M.G., 1996, Free-boundary model for local thermal coagulation. In: S.L. Jacques (Ed.) Laser – Tissue Interaction VII, Proceedings of SPIE, 2681, 81 – 91. [7] Lubashevsky, I.A., Priezzhev, A.V., Gafiychuk, V.V. and Cadjan, M.G., 1996, Dynamic free boundary model for laser thermal tissue ´ coagulation. In: H.J. Albrecht, G. Delacretaz, T.H. Meier, R.W. Steiner and L.O. Svaasand (Eds) Laser – Tissue Interaction and Tissue Optics II, Proceedings of SPIE, 2923, 48 – 57. [8] Lubashevsky, I.A., Priezzhev, A.V., Gafiychuk, V.V. and Cadjan, M.G., 1997, Local thermal coagulation due to laser – tissue interaction as irreversible phase transition. Journal of Biomedical Optics, 2, 95 – 105.

396

B. Y. Datsko et al.
[15] Weinbaum, S., Xu, L.X., Zhu, L. and Ekpene, A., 1997, A new fundamental bioheat equation for muscle tissue. Part I. Blood perfusion term. ASME Journal of Biomechanical Engineering, 119, 278 – 288. [16] Weinbaum, S. and Jiji, L.M., 1985, A new simplified bioheat equation for the effect of blood flow on local average tissue temperature. ASME Journal of Biomechanical Engineering, 107, 131 – 139. [17] Jacques, S.L., 1992, Laser – tissue interactions: Photochemical, photothermal, and photomechanical. Surgical Clinics of North America, 72, 531 – 558. [18] Welch, A.J., Torres, J.H. and Cheong, W.-F., 1989, Laser Physics and Laser – Tissue Interaction. Texas Heart Institute Journal, 16, 14, 1 – 9. [19] Robinson, D.S., Parel, J.-M., Denham, D.B., Gonzalez-Cirre, X., Manns, F., Milne, P., Schachner, R.D., Herron, A.J., Comander, J. and Hauptmann, G., 1998, Interstitial Laser Hyperthermia Model Development Minimally Invasive Therapy of Breast Carcinoma. Journal of the American College of Surgery, 186, 284 – 292. [20] Karaa, S., Zhang, J. and Yang F., 2005, A numerical study of a 3D bioheat transfer problem with different spatial heating. Mathematics and Computers in Simulation, 68, 375 – 388. [21] Dua, R. and Chakraborty, S., 2005, A novel modeling and simulation technique of photothermal interactions between lasers and living biological tissues undergoing multiple changes in phase. Computers in Biology and Medicine, 35, 447 – 462.

[9] Lubashevsky, I.A., Priezzhev, A.V. and Gafiychuk, V.V., 1997, Free boundary model for local thermal coagulation. Growth of a spherical and cylindrical necrosis domain. In: S.L. Jacques (Ed.) Laser – Tissue Interaction VIII, Proceedings of SPIE, 2975, 43 – 53. [10] Lubashevsky I.A. and Priezzhev, A.V., 1998, Laser induced heat diffusion limited tissue coagulation. I. Form of the necrosis boundary ´ caused by random temperature nonuniformities. In: G. Delacretaz, L.O. Svaasand, R.W. Steiner, R. Pini and G. Godlewski (Eds) Laser – Tissue Interaction, Tissue Optics, and Laser Welding III, Proceedings of SPIE, 3195, 143 – 150. [11] Lubashevsky, I.A., Priezzhev, A.V. and Gafiychuk, V.V., 1998, Effective interface dynamics of heat diffusion limited by thermal coagulation. Journal of Biomedical Optics, 3, 102 – 111. [12] Lubashevsky, I.A. and Priezzhev, A.V., 1998, Laser induced heat diffusion limited tissue coagulation. II. Effect of random temperature nonuniformities on the form of a spherical and cylindrical necrosis domain. In: S.L. Jacques (Ed.) Laser – Tissue Interaction IX, Proceedings of SPIE, 3254, 61 – 68. [13] Lubashevsky, I.A. and Priezzhev, A.V., 1999, Effect of the blood vessel discreteness on the necrosis formation during laser induced thermal coagulation limited by heat diffusion. Journal of Biomedical Optics, 4, 248 – 255. [14] Gafiychuk, V.V. and Lubashevsky, I.A., 1999, Mathematical Description of Heat Transfer in Living Tissue (Lviv: VNTL Publishers). Available online at: http://xxx.lanl.gov/abs/adap-org/9911001,9911002. (Accessed 30.12.2000).