1: %\documentclass[aps,prl,preprint,groupedaddress]{revtex4}

2: \documentclass[aps,pre,twocolumn,english,superscriptaddress,showpacs,floatfix]{revtex4}

3: %\documentclass[aps,prl,twocolumn,groupedaddress]{revtex4}

4: 

5: \usepackage{amsmath,bbm}

6: \usepackage{psfig,graphics,graphicx}

7: \usepackage{babel}

8: 

9: 

10: 

11: \begin{document}

12: 

13: % Use the \preprint command to place your local institutional report

14: % number in the upper righthand corner of the title page in preprint mode.

15: % Multiple \preprint commands are allowed.

16: % Use the 'preprintnumbers' class option to override journal defaults

17: % to display numbers if necessary

18: %\preprint{}

19: 

20: %Title of paper

21: \title{Topological constraints on

22: spiral wave dynamics in spherical geometries with inhomogeneous

23: excitability}

24: 

25: \author{J\"orn Davidsen}

26: \email[]{davidsen@mpipks-dresden.mpg.de}

27: \affiliation{Max-Planck-Institut f\"ur Physik Komplexer Systeme,

28: N\"othnitzer Strasse 38, 01187 Dresden, Germany}

29: \affiliation{Chemical Physics Theory Group, Department of

30: Chemistry, University of Toronto, Toronto, ON M5S 3H6, Canada}

31: \author{Leon Glass}

32: \email[]{glass@cnd.mcgill.ca}

33: \affiliation{Department of Physiology, McGill University,

34: Montreal, Quebec H3G 1Y6, Canada}

35: \author{Raymond Kapral}

36: \email[]{rkapral@chem.utoronto.ca} \affiliation{Max-Planck-Institut

37: f\"ur Physik Komplexer Systeme,

38: N\"othnitzer Strasse 38, 01187 Dresden, Germany}

39: \affiliation{Chemical Physics

40: Theory Group, Department of Chemistry, University of Toronto,

41: Toronto, ON M5S 3H6, Canada}

42: 

43: 

44: \date{\today}

45: 

46: \begin{abstract}

47: We analyze the way topological constraints and inhomogeneity in

48: the excitability influence the dynamics of spiral waves on spheres

49: and punctured spheres of excitable media. We generalize the

50: definition of an index such that it characterizes not only each

51: spiral but also each hole in punctured, oriented, compact,

52: two-dimensional differentiable manifolds and show that the sum of

53: the indices is conserved and zero. We also show that heterogeneity

54: and geometry are responsible for the formation of various spiral

55: wave attractors, in particular, pairs of spirals in which one

56: spiral acts as a source and a second as a sink --- the latter

57: similar to an antispiral. The results provide a basis for the

58: analysis of the propagation of waves in heterogeneous excitable

59: media in physical and biological systems.

60: 

61: 

62: 

63: 

64: 

65: \end{abstract}

66: 

67: % insert suggested PACS numbers in braces on next line

68: \pacs{87.19-j, 05.40.-a, 89.75.-k}

69: % insert suggested keywords - APS authors don't need to do this

70: %\keywords{}

71: 

72: %\maketitle must follow title, authors, abstract, \pacs, and \keywords

73: \maketitle

74: 

75: % body of paper here - Use proper section commands

76: % References should be done using the \cite, \ref, and \label commands

77: 

78: \section{Introduction}

79: Geometry and inhomogeneity influence pattern formation in chemical

80: and biological systems. \cite{murray89,kapral} One example where

81: these two factors play a crucial role is in the experimental

82: observations of distinctive spiral wave dynamics on the surfaces

83: of spherical beads, which are excitable inhomogeneous chemical

84: media \cite{maselko89,maselko90}. A biological example is the

85: origin of abnormal cardiac rhythms in the human heart which depend

86: on the anatomical substrate. The heart possesses a complex

87: nonplanar geometry with multiple chambers, with holes

88: corresponding to valves and blood vessels. Some serious

89: arrhythmias are associated with circulating spiral waves similar

90: to those observed in chemical media \cite{winfree01}. Since an

91: abnormal anatomical substrate is a common finding in patients with

92: some types of cardiac arrhythmias, and interventions that modify

93: the anatomy are an accepted form of therapy \cite{stein02},

94: theoretical analyses of the interplay between geometry of the

95: substrate and dynamics may help in the therapy of cardiac

96: arrhythmias.

97: 

98: In this paper we study spiral wave dynamics on (punctured) spheres

99: with spatially inhomogeneous excitability. We show for punctured

100: spheres that the sum of indices which characterize each spiral has

101: to be zero. Moreover, we demonstrate that topological constraints

102: imposed by the spherical geometry and inhomogeneity in

103: excitability lead to the formation of pairs of spirals, with

104: distinctive transient dynamics or as stable attractors, in which

105: one spiral acts as a source and a second as a sink leading to a

106: source-sink pair under a broad range of conditions. Our results

107: explain the experimental observations of spirals on spherical

108: beads \cite{maselko89,maselko90}.  While we do not consider

109: detailed models of cardiac wave propagation, our results may apply

110: to some generic aspects of atrial arrhythmias because the thin

111: walls of the atria can be described as two-dimensional (2d)

112: inhomogeneous excitable media with specific geometrical features.

113: \cite{3d}

114: 

115: 

116: \section{An Index Theorem for Phase Singularities}

117: 

118: The mathematical description of spiral waves is based on the

119: notion of phase which in turn allows one to characterize spiral

120: waves by an index. From this description, a number of topological

121: results placing restrictions on spiral wave dynamics can be

122: derived. \cite{glass77,winfree83,sumners87,winfree01,cruz-white03}

123: 

124: With the exception of a finite number of singular points, with

125: each point in an orientable and compact two-dimensional

126: differentiable manifold $M$ we identify a unique phase lying on

127: the unit circle, $\Phi \in S^1$. The resulting phase map or phase

128: field is assumed to be continuously differentiable, except at the

129: singular points. The manifold can be triangulated \cite{hocking88}

130: (subdividing it into a set of polygons), where none of the edges

131: or vertices of the polygons pass through a singularity. The index,

132: $I$ (sometimes also called the topological charge or winding

133: number) of a curve $C$ bounding a polygon is found by computing

134: the line integral

135: \begin{equation}\label{top_charge}

136:     2 \pi I = \oint_C \nabla \Phi \cdot d{\bf l},

137: \end{equation}

138: where the polygon is always traversed in a clockwise orientation.

139: By continuity of $\nabla \Phi$, $I$ must be an integer. The index

140: of a singular point is uniquely defined as the index of any curve

141: $C$ provided that $C$ encircles the point but no other singular

142: points. The index of a curve that does not enclose any singular

143: points is obviously zero.

144: 

145: If the manifold $M$ has no boundaries, each edge of the

146: triangulation is an edge of two polygons. Since the line integral

147: adds up the change in phase along the various edges of the

148: polygon, the sum of the indices of the singular points for a phase

149: field in $M$ is

150: \begin{equation}\label{indextheorem}

151: \Sigma_M I =0,

152: \end{equation}

153: where the sum is over all the singular points.  This follows since

154: the contribution of the change in phase of each edge to the total

155: integral is counted twice, but since the edge is traversed in

156: opposite directions each time, the net contribution of each edge

157: is zero. This index theorem is applicable to tori, and other

158: surfaces of genus different from 0. However, unlike the more

159: familiar Poincar\'e index theorem (see Ref. \cite{guillemin74} p.

160: 74 for vector fields) the sum of the indices of the singular

161: points does not depend on the genus of the surface.

162: 

163: This index theorem for manifolds without boundaries can be

164: extended to manifolds with boundaries. In the following, we will

165: consider the case of structures that arise from puncturing

166: orientable and compact two-dimensional differentiable manifolds.

167: The index of a hole can be uniquely defined as the index of a

168: curve $C$ provided that $C$ encircles the hole but no other holes

169: or singular points and $C$ is positively oriented with respect to

170: the domain which contains the hole and is bounded by $C$. Applying

171: this definition and taking the summation in Eq.

172: (\ref{indextheorem}) over the singularities and the hole, or the

173: holes if there is more than one hole, the index theorem can be

174: proven by the same line of arguments as for the case of manifolds

175: without boundaries.

176: 

177: This extension is important in the heart, for example, where the

178: atrium is punctured by valves and veins. In such cases one is led

179: to consider manifolds with holes; for example, a sphere with a

180: hole. A sphere with a hole is topologically equivalent to a disk,

181: and, indeed, the results for disks and for spheres with holes are

182: consistent: For the disk $D^2$, bounded by a curve $C$,

183: \begin{equation}\label{dindextheorem}

184: \Sigma_{D^2} I =\oint_C \nabla \Phi \cdot d{\bf l},

185: \end{equation}

186: so that the sum of the indices of the singular points in the disk

187: is equal to the index of the curve $C$ bounding the disk. If there

188: is a single singular point on the disk, with an index of +1, the

189: index of the curve bounding the disk will also be +1. Imagine now

190: the boundary of the disk to be brought together (like a

191: draw-string bag) so that the boundary of the disk now defines a

192: hole in the sphere. In this geometry the curve $C$ will be

193: traversed in an opposite orientation (the hole is now inside $C$)

194: from the direction it was traversed when it was the boundary of

195: the disk. Now if there is a singular point with an index of +1 on

196: the sphere, the index of the hole is -1, so that the sum of the

197: indices is again zero.

198: 

199: Since it is necessary to conserve the sum of the indices,

200: singularities of index $\pm 1$ usually arise and are destroyed in

201: pairs of opposite sign \cite{winfree83}. An exception occurs when

202: singularities are destroyed by collision into a boundary, so that

203: the index of the singular point and the index of the boundary both

204: change simultaneously. Another exception occurs if there are

205: singularities with index different from one. In such cases

206: interactions between different singularities can lead to

207: destruction or creation of odd numbers of singularities

208: \cite{zaritski02}.

209: 

210: 

211: 

212: \section{The FitzHugh-Nagumo Equation}

213: 

214: The FitzHugh-Nagumo (FHN) equation, \cite{numerics}

215: \begin{eqnarray}

216: \frac{\partial u}{\partial t} &=& \epsilon^{-1} \left( -

217: \frac{u^3}{3} + u - v \right) + D_u \nabla^2 u,\nonumber\\

218: \frac{\partial v}{\partial t} &=& \epsilon \left( u - \alpha v +

219: \beta \right) + D_v \nabla^2 v,\label{equdyn}

220: \end{eqnarray}

221: where $u({\bf r},t)$ and $v({\bf r},t)$ are two scalar fields,

222: $\epsilon^2$ is the ratio of the time scales associated with the

223: two fields, and $D_u$ and $D_v$ are the constant diffusion

224: coefficients, is a

225: prototypical model for excitable media.

226: We choose $D_u = 2$ and $D_v = 0$. The real

227: parameters $\alpha$ and $\beta$ characterize the local dynamics

228: and, hence, the local excitability. Whenever $0<\alpha<1$, $\alpha

229: \epsilon^2 < 1$ and $|\beta| > \beta_H \equiv (1 - \alpha

230: \epsilon^2)^{1/2} (1/3 (2 \alpha + \alpha^2 \epsilon^2) - 1)$, the

231: FHN system is excitable. At $\beta_H$ a Hopf bifurcation occurs

232: such that for $|\beta| < \beta_H$ the system exhibits

233: oscillations. In the following we take $\alpha = 0.2$, $\epsilon=0.2$ and

234: $\beta > \beta_H = 0.863\dots$.

235: 

236: We consider a spherical shell whose outer and inner radii are

237: $R_e$ and $R_i$, respectively, and focus on thin spherical shells

238: where $R_i=40$, $R_e=42$. The radii are large enough to avoid a

239: curvature-dependent loss of excitability \cite{davydov02}, and the

240: shell is sufficiently thin that the dynamics is effectively 2d

241: corresponding to the dynamics on a sphere. \cite{shell} The

242: initial condition is a domain of ``excited" state, adjacent to a

243: domain of the ``refractory" state. Both domains have the forms of

244: slices of the same size oriented from the north to the south pole

245: \cite{details} and yield a pair of counter-rotating spirals.

246: 

247: In order to apply the topological results to the FHN equation, it

248: is necessary to first define the phase. We define a phase,

249: $\Phi({\bf r},t)$, based on the equation $\tan{\Phi({\bf r},t)} =

250: v({\bf r},t)/u({\bf r},t)$ if $v({\bf r},t)\neq 0$ and $u({\bf

251: r},t)\neq 0$. Thus, singular points at given $t$ are points ${\bf

252: r}$ in the medium for which $v({\bf r},t)=0$ and $u({\bf r},t)=0$.

253: For each $t$, we obtain a continuously differentiable phase map

254: ${\cal M}^t=\Phi({\bf :},t)|_{{\cal D}^t}$ that associates to each

255: point in a well-defined domain ${\cal D}^t$ a phase lying on the

256: unit circle, $\Phi \in S^1$. For our FHN medium, the domain is the

257: surface of a (punctured) sphere reduced by a finite number of

258: points where the phase is singular at fixed $t$.

259: 

260: Rotating spiral waves in the FHN equations are obviously

261: associated with a singular point which is called the spiral core.

262: In what follows, we assume that there are only single-armed

263: spirals so that a clockwise rotating source has an index of +1,

264: and a counterclockwise rotating source has index -1. A clockwise

265: rotating sink has an index of -1, and a counterclockwise rotating

266: sink has index +1. From Eq.~(\ref{indextheorem}), it is impossible

267: to have a single rotating spiral wave on a sphere; in addition,

268: there must be at least another singular point or a hole with

269: nonzero index.

270: 

271: For excitable media, a non-zero index of a hole implies that wave

272: fronts travel permanently around the hole such that the numbers of

273: fronts travelling clockwise and counterclockwise are different.

274: This includes the particularly simple case of a single wave front

275: travelling around the hole which can be considered as a spiral

276: wave associated with the hole.

277: 

278: \section{Spiral Wave Dynamics in Spherical Geometries}

279: \subsection{Dynamics in excitability gradients}

280: First consider homogeneous FHN media with a constant

281: $\beta=\beta_{ex}\equiv0.9$. The wave fronts emanating from the

282: spiral cores with opposite index ``annihilate" along the equator

283: such that each spiral determines the dynamics on half of the

284: sphere (see Fig.~\ref{exc_hom}, left panel) --- similar to what

285: has been observed in Ref.~\cite{yagisita98} for a different

286: excitable system.

287:  \begin{figure}

288:  \includegraphics[width=0.75\columnwidth]{comb}

289:  \caption{\label{exc_hom}

290: (Color online) Left: Spiral waves of excitation (light fronts) on

291: the sphere for constant $\beta=\beta_{ex}$ emanating from spiral

292: cores close to the poles on an equator projection. The white

293: arrows show the direction of propagation. One annihilation front

294: along the equator can be identified. Right: Sketch of the constant

295: gradient in the inhomogeneous case. The dashed lines are the

296: equi-$\beta$ lines and we choose $\beta_{max}=1.0$ and

297: $\beta_{min}=\beta_{ex}$. The angle $\phi$ describes the

298: orientation with respect to the axis from pole to pole. The

299: results described in the text do not depend qualitatively on the

300: choice of $\phi$ or $\beta_{min}$ and $\beta_{max}$ as long as

301: they yield stable spirals.}

302:  \end{figure}

303: We have shown that this behavior is robust with respect to

304: disorder in excitability with small amplitude and correlation

305: length. If random, uncorrelated spatial variations in $\beta$

306: exist on length scales much smaller than the diameter of the

307: spiral core meander \cite{random_details}, the dynamics is able to

308: average over such small-scale inhomogeneities. The robustness

309: explains why such states have been experimentally observed in some

310: chemical reactions on spherical beads which are intrinsically

311: inhomogeneous \cite{maselko90}.

312: 

313: Applying a constant gradient in $\beta$ as sketched in the right

314: panel of Fig.~\ref{exc_hom} leads to a different scenario. The time

315: evolution of the spiral pair may be partitioned into four distinct

316: regimes as shown in Figs.~\ref{beta} and \ref{d}.

317:  \begin{figure}

318:  \includegraphics[width=\columnwidth]{beta_ang089_trans_new}

319:  \caption{\label{beta}

320: (Color online) The local excitability $\beta({\bf r}(t))$ at the

321: spiral cores versus time. Gradient-induced motion of the two

322: spiral cores leads to a change in the local excitability at the

323: cores with time. The spiral period in the final state is

324: $T_0=13.2\pm0.1$. Four different regimes can be identified (see

325: text). Inset: The final bound pair of counter-rotating spirals in

326: regime III for $\phi=51.0^\circ$ is shown on an equator projection

327: such that the point of lowest excitability lies on the central

328: longitude. The spiral closer to the equator has index $-1$ while

329: the other one has index $+1$.}

330:  \end{figure}

331:  \begin{figure}

332:  \includegraphics[width=\columnwidth]{d_ang089_trans}

333:  \caption{\label{d}

334: (Color online) The distance between the two spiral cores $d(t)$

335: versus $t$. Gradient-induced motion of the two spiral cores leads

336: to a change in the distance $d$ between the cores with time. The

337: lower and upper curves correspond to the distance in

338: $\mathbbm{R}^3$ and $S^2$, respectively. The dashed lines are the

339: respective upper bounds given by the size of the sphere.}

340:  \end{figure}

341: Because of the gradient, the frequencies of the two spirals differ

342: since a higher value of $\beta$ corresponds to lower excitability,

343: which generally implies a lower spiral frequency \cite{winfree91}.

344: During a short transient T, the spiral with the higher frequency

345: assumes control of the dynamics \cite{krinsky83} on the sphere.

346: The location on the sphere, where wave fronts emanating from the

347: two spiral cores annihilate, moves toward the core of the

348: low-frequency spiral. Finally, the low-frequency spiral core is

349: pushed farther from the high-frequency spiral core

350: \cite{krinsky83,ermakova86}(see Fig.~\ref{d}). After this short

351: transient, the wave fronts travel from pole to pole leading to the

352: creation of a source-sink pair. This (intermediate) state is shown

353: in Fig.~\ref{anti_series} and corresponds to regime I in

354: Figs.~\ref{beta} and \ref{d}. Viewed from the low excitability end

355: of the sphere, the waves wind into a small region about the core,

356: reminiscent of the structure of antispirals, i.e., inward moving

357: spirals seen in oscillatory media \cite{gong03}. However, the

358: origin of this inward spiral motion in oscillatory media differs

359: and is distinct from that observed here. In oscillatory media,

360: either spirals or antispirals are stable depending on system

361: parameters and the wave length diverges on the border in the

362: parameter space between these two regimes. Thus, antispirals exist

363: independently of spirals. This is not the case here because the

364: generation of an inward moving spiral relies on the presence of a

365: spiral source and spherical geometry. For example, consider the

366: FHN system with a disc geometry and a radial gradient in

367: excitability such that the maximum value of $\beta$ is located in

368: the center of the disc. In this case, a source-sink pair cannot

369: occur because the high-frequency spiral, acting as a source, would

370: push the other (low-frequency) spiral out of the system, excluding

371: the presence of any strong random inhomogeneities in excitability

372: which may pin the low-frequency spiral and prevent its motion

373: (see, e.g., \cite{biktashev01}). The lower panel of

374: Fig.~\ref{anti_series} shows that remnants of the low frequency

375: spiral persist in a small area around the core. It has been

376: speculated that antispiral waves might occur in cardiac tissue

377: \cite{vanag01}. While excitable media cannot support a regime of

378: exclusive antispirals due to their excitable character, our

379: results show that source-sink pairs with similar characteristics

380: could form in the heart where the underlying dynamics is

381: excitable, the medium is inhomogeneous and the topology is similar

382: to a (punctured) sphere.

383: 

384: 

385: Spiral dynamics of the type described above has been observed by

386: Maselko and Showalter \cite{maselko89} in experiments on the

387: excitable Belouzov-Zhabotinsky reaction on spherical beads. They

388: attributed the generation of spiral source-sink pairs to

389: inhomogeneities in the medium related to differing chemical

390: environments. This is consistent with our findings for systems

391: with gradients and is further confirmed by the work reported in

392: Ref.~\cite{yagisita98}. While the generation of source-sink pairs

393: due to a gradient in the FHN medium investigated here is only an

394: intermediate state (but one that persists for approximately 240

395: spiral periods in our simulations), random spatial variations of

396: the excitability with a correlation length comparable to the

397: diameter of the spiral core meander or larger can lead to a final

398: state consisting of such a source-sink pair \cite{random_details}.

399: This is due to the fact that the source can be trapped in a region

400: of depressed local excitability.

401:  \begin{figure}

402:  \includegraphics[width=\columnwidth]{anti_series}

403:  \caption{\label{anti_series}

404: (Color online) Waves of excitation on the sphere in regime I of

405: the gradient-induced dynamics shown in Figs.~\ref{beta} and

406: \ref{d}. A source-sink pair has formed. For random spatial

407: variations of excitability with a correlation length comparable to

408: the diameter of the spiral core meander, the final state is very

409: similar to the one shown here \cite{random_details}. Upper panel

410: from left to right: View centered at the north pole, south pole

411: and the equator. The source at the south pole has index $-1$ and

412: the sink at the north pole index $+1$. The black circles show

413: possible choices of $C$. The white arrow shows the direction of

414: wave propagation. Lower panel: Dynamics at the north pole. Time

415: increases from left to right with $\Delta t = 2.5$ between

416: snapshots.}

417:  \end{figure}

418: 

419: Not only does the gradient in the FHN medium change the local

420: excitability but it also induces a drift of the spiral cores

421: \cite{biktashev95}. For our model, the drift is rather slow

422: compared to the transition to the source-sink pair which takes

423: place during the transient regime T. This can be seen in

424: Fig.~\ref{beta}. (The fluctuations in $\beta({\bf r}(t))$ are due

425: to spiral meandering.) In regime I, the dominating spiral drifts

426: toward lower excitability and its wave fronts continuously push

427: the other core in the opposite direction, thus, keeping the

428: distance $d$ between the cores approximately constant (see

429: Fig.~\ref{d}). The fluctuations in $d(t)$ are again due to

430: meandering of the spiral cores. Because of this drift, the local

431: excitabilities at the two spiral cores approach each other until

432: they become equal.

433: 

434: At this point, regime II in Figs.~\ref{beta} and \ref{d} is

435: entered. The dynamics change drastically: the slaved spiral

436: reverses its drift direction and regains control over its own

437: dynamics. Both spirals drift toward lower excitability. Due to the

438: geometric constraints imposed by the spherical geometry, the

439: spirals approach each other until they form a bound pair (at $t

440: \approx 7500$).

441: 

442: They finally reach a stable state (at $t \approx 9500$)

443: corresponding to regime III in Figs.~\ref{beta} and \ref{d}.

444: Neither the \emph{average} distance between the spirals nor the

445: \emph{average} local excitability changes further. Yet, on top of

446: the persisting unsynchronized meandering of the two spiral cores,

447: the bound pair slowly moves along a (closed) equi-$\beta$ curve on

448: the surface of the sphere. The direction of the motion depends on

449: the initial condition, i.e., whether the spiral with positive

450: index was initially closer to the point of lowest excitability or

451: to the point of highest excitability than the spiral with negative

452: index. The wave dynamics generated by this bound pair is shown in

453: the inset of Fig.~\ref{beta}.

454: 

455: While kinematic theory applies only to spirals with large cores,

456: it is instructive to note that this theory predicts that the

457: direction of the drift due to gradients depends on the model

458: system and its parameters \cite{mikhailov94}. Although our

459: simulations have shown the same direction of the drift for a range

460: of parameters in the meander region of the FHN phase diagram, it

461: is conceivable that, under different circumstances favoring a

462: drift toward higher excitability, one spiral could act as a

463: permanent source and a source-sink pair could be the final

464: attractor. Such a scenario would also be consistent with the

465: experimental findings in Ref. \cite{maselko89}.

466: 

467: \subsection{Punctured spheres}

468: Next, we consider a homogeneously excitable sphere with a single

469: hole. Two scenarios can be observed depending on the location of

470: the hole with respect to the spiral pair. If a spiral wave is not

471: permanently attached to the hole, the dynamics is very similar to

472: the case without any hole. If one of the spirals is permanently

473: attached to the hole, the frequency of this spiral is lowered. The

474: size of the hole determines the frequency of the spiral because

475: the wave front has to travel around the hole. The transient

476: dynamics is similar to that in regime T for the case with a

477: gradient; however, no drift of the spiral cores is induced and the

478: final state is a spiral source-sink pair as shown in

479: Fig.~\ref{series_hole}. Not only is the net index conserved during

480: the transition to a spiral source-sink pair but so is the index of

481: the individual spirals: during the transition an outgoing

482: counterclockwise (clockwise) oriented spiral is converted into an

483: ingoing clockwise (counterclockwise) oriented spiral. Thus, the

484: formation of spiral source-sink pairs conforms with the

485: topological constraints.

486: 

487: If a gradient as well as a hole is present, the spiral drift

488: discussed earlier also determines the final state, which depends

489: on the hole's location in the gradient field. For instance, if the

490: point of lowest excitability in the medium is on the hole

491: boundary, simulations show that the spiral which is not attached

492: to the hole will stabilize close to this point. In this final

493: state odd numbers of wavefronts are attached to the hole, again

494: conserving the topological charge of the hole.

495: 

496: 

497: \begin{figure}

498:   \includegraphics[width=\columnwidth]{series_hole}

499:   \caption{\label{series_hole}

500: (Color online) Waves of excitation on the punctured sphere

501: propagating towards the hole. A source-sink spiral wave pair has

502: formed in the final state. Upper panel from left to right: View

503: centered at the north pole, south pole and the equator. Lower

504: panel: Dynamics at the south pole. Time increases from left to

505: right with $\Delta t = 2.5$ between snapshots.}

506: \end{figure}

507: 

508: \section{Concluding Remarks}

509: Inhomogeneities due to spatially-varying excitability on

510: (punctured) spherical shells lead to complex spiral wave dynamics

511: and the formation of source-sink spiral pairs in excitable media.

512: The results presented here are immediately applicable to excitable

513: media in more complicated geometries such as tori or multi-holed

514: tori and to situations in which multi-armed spirals are found.

515: This includes mathematical modelling of cardiac tissue. The

516: approach taken in this paper stresses constraints and aspects that

517: apply to, and must be observed in, all realistic models of the

518: heart satisfying certain criteria of continuity. There are also

519: implications for the treatment of cardiac arrhythmias. In

520: cardiology it is sometimes possible to develop maps showing the

521: timing of the excitation over limited regions of heart

522: \cite{stein02}. In this case, a sink might be confused for a

523: source (of the arrhythmia), and this might have implications for

524: the diagnosis of the mechanism and the choice of therapy. The

525: current work shows how partial knowledge about what is happening

526: in some regions that could be observed, might be helpful in

527: establishing properties of dynamics that could not be observed.

528: While the types of sinks we have described here have only been

529: observed in chemical media \cite{maselko89,maselko90} so far, we

530: certainly expect their existence in the cardiological domain.

531: 

532: 

533: We thank F. Chavez, M. B\"ar, S. G. Whittington and D. Sumners for

534: helpful discussions and G. Rousseau for providing numerical tools.

535: This work was supported in part by a grant from MITACS.

536: 

537: % Specify following sections are appendices. Use \appendix* if there

538: % only one appendix.

539: %\appendix

540: %\section{}

541: 

542: % If you have acknowledgments, this puts in the proper section head.

543: %\begin{acknowledgments}

544: % put your acknowledgments here.

545: %\end{acknowledgments}

546: 

547: % Create the reference section using BibTeX:

548: \begin{thebibliography}{26}

549: \expandafter\ifx\csname

550: natexlab\endcsname\relax\def\natexlab#1{#1}\fi

551: \expandafter\ifx\csname bibnamefont\endcsname\relax

552:   \def\bibnamefont#1{#1}\fi

553: \expandafter\ifx\csname bibfnamefont\endcsname\relax

554:   \def\bibfnamefont#1{#1}\fi

555: \expandafter\ifx\csname citenamefont\endcsname\relax

556:   \def\citenamefont#1{#1}\fi

557: \expandafter\ifx\csname url\endcsname\relax

558:   \def\url#1{\texttt{#1}}\fi

559: \expandafter\ifx\csname

560: urlprefix\endcsname\relax\def\urlprefix{URL }\fi

561: \providecommand{\bibinfo}[2]{#2}

562: \providecommand{\eprint}[2][]{\url{#2}}

563: 

564: 

565: \bibitem[{\citenamefont{Murray}(1989)}]{murray89}

566: \bibinfo{author}{\bibfnamefont{J.~D.} \bibnamefont{Murray}},

567:   \emph{\bibinfo{title}{Mathematical Biology}}

568:   (\bibinfo{publisher}{Springer-Verlag}, \bibinfo{address}{New York},

569:   \bibinfo{year}{1989});

570:   \bibinfo{note}{and references herein.}

571: 

572: \bibitem[{\citenamefont{Kapral and Showalter}(1996)}]{kapral}

573: \bibinfo{editor}{\bibfnamefont{R.}~\bibnamefont{Kapral}} \bibnamefont{and}

574:   \bibinfo{editor}{\bibfnamefont{K.}~\bibnamefont{Showalter}}, eds.,

575:   \emph{\bibinfo{title}{Chemical Waves and Patterns}}

576:   (\bibinfo{publisher}{Kluwer Academic Publishers},

577:   \bibinfo{address}{Dordrecht, Netherlands},

578:   \bibinfo{year}{1996});   \bibinfo{note}{and references herein.}

579: 

580: \bibitem[{\citenamefont{Maselko and Showalter}(1989)}]{maselko89}

581: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Maselko}} \bibnamefont{and}

582:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Showalter}},

583:   \bibinfo{journal}{Nature (London)} \textbf{\bibinfo{volume}{339}},

584:   \bibinfo{pages}{609} (\bibinfo{year}{1989}).

585: 

586: \bibitem[{\citenamefont{Maselko and Showalter}(1990)}]{maselko90}

587: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Maselko}} \bibnamefont{and}

588:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Showalter}},

589:   \bibinfo{journal}{React. Kinet. Catal. Lett.} \textbf{\bibinfo{volume}{42}},

590:   \bibinfo{pages}{263} (\bibinfo{year}{1990}).

591: 

592: \bibitem[{\citenamefont{Winfree}(2001)}]{winfree01}

593: \bibinfo{author}{\bibfnamefont{A.~T.} \bibnamefont{Winfree}},

594:   \emph{\bibinfo{title}{The Geometry of Biological Time}}

595:   (\bibinfo{publisher}{Springer-Verlag}, \bibinfo{address}{New York},

596:   \bibinfo{year}{2001}).

597: 

598: \bibitem[{\citenamefont{Stein et~al.}(2002)\citenamefont{Stein, Markowitz,

599:   Mittal, Slotwiner, Iwai, and Lerman}}]{stein02}

600: \bibinfo{author}{\bibfnamefont{K.~M.} \bibnamefont{Stein}},

601:   \bibinfo{author}{\bibfnamefont{S.~M.} \bibnamefont{Markowitz}},

602:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mittal}},

603:   \bibinfo{author}{\bibfnamefont{D.~J.} \bibnamefont{Slotwiner}},

604:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Iwai}}, \bibnamefont{and}

605:   \bibinfo{author}{\bibfnamefont{B.~B.} \bibnamefont{Lerman}},

606:   \bibinfo{journal}{Chaos} \textbf{\bibinfo{volume}{12}}, \bibinfo{pages}{740}

607:   (\bibinfo{year}{2002}).

608: 

609: \bibitem[{num()}]{3d}

610: \bibinfo{note}{Some aspects of cardiac arrhythmias may arise from

611: scroll wave filament instabilities which depend on the 3d nature

612: of cardiac tissue \cite{winfree87,mikhailov03}.}

613: 

614: \bibitem[{\citenamefont{Glass}(1977)}]{glass77}

615: \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Glass}},

616:   \bibinfo{journal}{Science} \textbf{\bibinfo{volume}{198}},

617:   \bibinfo{pages}{321} (\bibinfo{year}{1977}).

618: 

619: \bibitem[{\citenamefont{Winfree and Strogatz}(1983)}]{winfree83}

620: \bibinfo{author}{\bibfnamefont{A.~T.}~\bibnamefont{Winfree}}

621: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{S.~H.} \bibnamefont{Strogatz}},

622:   \bibinfo{journal}{Physica D} \textbf{\bibinfo{volume}{8}},

623:   \bibinfo{pages}{35} (\bibinfo{year}{1983}).

624: 

625: \bibitem[{\citenamefont{Sumners}(1987)}]{sumners87}

626: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Sumners}},

627: \bibinfo{title}{Knots, macromolecules and chemical dynamics},

628: \bibnamefont{in}  \emph{\bibinfo{title}{Graph Theory and

629: Topology in Chemistry}},

630: \bibinfo{editor}{\bibfnamefont{R.~B.}~\bibnamefont{King}} \bibnamefont{and}

631: \bibinfo{editor}{\bibfnamefont{D.}~\bibnamefont{Rouvray}}, eds.,

632: (\bibinfo{publisher}{Elsevier},

633: \bibinfo{address}{Netherlands},

634: \bibinfo{year}{1987}), p. 3.

635: 

636: \bibitem[{\citenamefont{Cruz-White}(2003)}]{cruz-white03}

637: \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Cruz-White}},

638: \bibnamefont{in}  \emph{\bibinfo{title}{Topology of Spiral Waves

639: in Excitable Media}}, (\bibinfo{publisher}{Ph.D. Thesis},

640: \bibinfo{address}{Florida State University Department of Mathematics},

641: \bibinfo{year}{2003}).

642: 

643: 

644: \bibitem[{\citenamefont{Hocking}(1988)}]{hocking88}

645: \bibinfo{author}{\bibfnamefont{J.~G.} \bibnamefont{Hocking}}

646: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{G.~S.} \bibnamefont{Young}},

647:   \emph{\bibinfo{title}{Topology}}

648:   (\bibinfo{publisher}{Dover}, \bibinfo{address}{New York},

649:   \bibinfo{year}{1988}).

650: 

651: 

652: \bibitem[{\citenamefont{Guillemin and Pollack}(1974)}]{guillemin74}

653: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Guillemin}},

654: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Pollack}},

655:   \bibinfo{title}{Differential Topology}

656: (\bibinfo{publisher}{Prentice Hall}, \bibinfo{address}{Englewood Cliffs, NJ},

657:   \bibinfo{year}{1974}).

658: 

659: 

660: \bibitem[{\citenamefont{Winfree}(1987)}]{winfree87}

661: \bibinfo{author}{\bibfnamefont{A.~T.} \bibnamefont{Winfree}},

662:   \emph{\bibinfo{title}{When Time Breaks Down}}

663:   (\bibinfo{publisher}{Princeton University Press},

664: \bibinfo{address}{Princeton, NJ},

665:   \bibinfo{year}{1987}).

666: 

667: \bibitem[{\citenamefont{Alonso et~al. }(2003)}]{mikhailov03}

668: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Alonso}},

669: \bibinfo{author}{\bibfnamefont{F.} \bibnamefont{Sagu\'es}} \bibnamefont{and}

670:   \bibinfo{author}{\bibfnamefont{A.~S.}~\bibnamefont{Mikhailov}},

671:   \bibinfo{journal}{Science} \textbf{\bibinfo{volume}{299}},

672:   \bibinfo{pages}{1722} (\bibinfo{year}{2003}).

673: 

674: \bibitem[{\citenamefont{Zaritski and Pertsov}(2002)}]{zaritski02}

675: \bibinfo{author}{\bibfnamefont{R.~M.} \bibnamefont{Zaritski}} \bibnamefont{and}

676:   \bibinfo{author}{\bibfnamefont{A.~M.}~\bibnamefont{Pertsov}},

677:   \bibinfo{journal}{Phys. Rev. E} \textbf{\bibinfo{volume}{66}},

678:   \bibinfo{pages}{066120} (\bibinfo{year}{2002}).

679: 

680: 

681: \bibitem[{num()}]{numerics}

682: \bibinfo{note}{We solve Eq. (\ref{equdyn}) numerically using an algorithm that

683:   automatically adjusts the time step to achieve an effcient simulation while

684:   controling the error in the solution \cite{rousseau00}.}

685: 

686: \bibitem[{\citenamefont{Davydov et~al.}(2002)\citenamefont{Davydov, Manz,

687:   Steinbock, and M{\"u}ller}}]{davydov02}

688: \bibinfo{author}{\bibfnamefont{V.~A.} \bibnamefont{Davydov}},

689:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Manz}},

690:   \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Steinbock}},

691:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{S.~C.}

692:   \bibnamefont{M{\"u}ller}}, \bibinfo{journal}{Europhys. Lett.}

693:   \textbf{\bibinfo{volume}{59}}, \bibinfo{pages}{344} (\bibinfo{year}{2002}).

694: 

695: \bibitem[{det()}]{shell}

696: \bibinfo{note}{It is computationally convenient to carry out the calculations

697: in a thin spherical shell constructed from a cubic rectangular

698: grid. The shell thickness is small compared to characteristic

699: wavelengths and diffusion lengths so that the small shell

700: thickness does not influence the spiral wave dynamics.}

701: 

702: \bibitem[{det()}]{details}

703: \bibinfo{note}{The initial conditions are as in Ref.~\cite{chavez01} with

704:   spherical coordinates $\Delta\theta=\pi/2$ and $\Delta\phi=\pi$.}

705: 

706: \bibitem[{\citenamefont{Yagisita et~al.}(1998)\citenamefont{Yagisita, Mimura,

707:   and Yamada}}]{yagisita98}

708: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Yagisita}},

709:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Mimura}}, \bibnamefont{and}

710:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Yamada}},

711:   \bibinfo{journal}{Physica D} \textbf{\bibinfo{volume}{124}},

712:   \bibinfo{pages}{126} (\bibinfo{year}{1998}).

713: 

714: \bibitem[{ran()}]{random_details}

715: \bibinfo{note}{For a Gaussian distribution of $\beta$ in boxes of volume

716:   $0.5^3$ with mean 0.95 and standard deviation 0.013, the diameter of the

717:   spiral core meander is roughly 10 space units.}

718: 

719: \bibitem[{\citenamefont{Winfree}(1991)}]{winfree91}

720: \bibinfo{author}{\bibfnamefont{A.~T.} \bibnamefont{Winfree}},

721:   \bibinfo{journal}{Chaos} \textbf{\bibinfo{volume}{1}}, \bibinfo{pages}{303}

722:   (\bibinfo{year}{1991}).

723: 

724: %\bibitem[{\citenamefont{Gelfand and Tsetlin}(1960)}]{gelfand60}

725: %\bibinfo{author}{\bibfnamefont{I. M.}~\bibnamefont{Gelfand}} \bibnamefont{and}

726: %  \bibinfo{author}{\bibfnamefont{M. L.}~\bibnamefont{Tsetlin}},

727: %  \bibinfo{journal}{Proc. Acad. Sci. USSR} \textbf{\bibinfo{volume}{131}},

728: %  \bibinfo{pages}{1242} (\bibinfo{year}{1960}).

729: 

730: \bibitem[{\citenamefont{Krinsky and Agladze}(1983)}]{krinsky83}

731: \bibinfo{author}{\bibfnamefont{V. I.}~\bibnamefont{Krinsky}} \bibnamefont{and}

732:   \bibinfo{author}{\bibfnamefont{K. I.}~\bibnamefont{Agladze}},

733:   \bibinfo{journal}{Physica D} \textbf{\bibinfo{volume}{8}},

734:   \bibinfo{pages}{50} (\bibinfo{year}{1983}).

735: 

736: \bibitem[{\citenamefont{Ermakova et al.}(1960)}]{ermakova86}

737: \bibinfo{author}{\bibfnamefont{E. A.}~\bibnamefont{Ermakova}},

738: \bibinfo{author}{\bibfnamefont{V. I.}~\bibnamefont{Krinsky}},

739: \bibinfo{author}{\bibfnamefont{A. V.}~\bibnamefont{Panfilov}},

740:  \bibnamefont{and}

741:   \bibinfo{author}{\bibfnamefont{A. M.}~\bibnamefont{Pertsov}},

742:   \bibinfo{journal}{Biofizika} \textbf{\bibinfo{volume}{31}},

743:   \bibinfo{pages}{318} (\bibinfo{year}{1986}).

744: 

745: 

746: \bibitem[{\citenamefont{Gong and Christini}(2003)}]{gong03}

747: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Gong}} \bibnamefont{and}

748:   \bibinfo{author}{\bibfnamefont{D.~J.} \bibnamefont{Christini}},

749:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{90}},

750:   \bibinfo{pages}{088302} (\bibinfo{year}{2003});

751: \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Brusch}},

752:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Nicola}}, \bibnamefont{and}

753:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{B{\"a}r}},

754:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{92}},

755:   \bibinfo{pages}{089801} (\bibinfo{year}{2004}).

756: 

757: \bibitem[{\citenamefont{Biktashev et~al.}(2001)}]{biktashev01}

758:   \bibinfo{author}{\bibfnamefont{V.~N.} \bibnamefont{Biktashev}}

759:   \bibinfo{author}{\bibfnamefont{A.~V.} \bibnamefont{Holden}},

760:   \bibinfo{author}{\bibfnamefont{S.~F.} \bibnamefont{Mironov}},

761:   \bibinfo{author}{\bibfnamefont{A.~M.} \bibnamefont{Pertsov}},

762: \bibnamefont{and}

763:   \bibinfo{author}{\bibfnamefont{A.~V.} \bibnamefont{Zaitsev}},

764:   \bibinfo{journal}{Int. J. Bifurcation Chaos Appl. Sci. Eng.}

765:   \textbf{\bibinfo{volume}{11}},

766:   \bibinfo{pages}{1035} (\bibinfo{year}{2001}).

767: 

768: \bibitem[{\citenamefont{Vanag and Epstein}(2001)}]{vanag01}

769: \bibinfo{author}{\bibfnamefont{V.~K.} \bibnamefont{Vanag}} \bibnamefont{and}

770:   \bibinfo{author}{\bibfnamefont{I.~R.} \bibnamefont{Epstein}},

771:   \bibinfo{journal}{Science} \textbf{\bibinfo{volume}{294}},

772:   \bibinfo{pages}{835} (\bibinfo{year}{2001}).

773: 

774: \bibitem[{\citenamefont{Biktashev and Holden}(1995)}]{biktashev95}

775: \bibinfo{author}{\bibfnamefont{V.~N.} \bibnamefont{Biktashev}}

776: \bibnamefont{and}

777:   \bibinfo{author}{\bibfnamefont{A.~V.} \bibnamefont{Holden}},

778:   \bibinfo{journal}{Chaos, Sol. Fract.} \textbf{\bibinfo{volume}{5}},

779:   \bibinfo{pages}{575} (\bibinfo{year}{1995}).

780: 

781: \bibitem[{\citenamefont{Mikhailov et~al.}(1994)\citenamefont{Mikhailov,

782:   Davydov, and Zykov}}]{mikhailov94}

783: \bibinfo{author}{\bibfnamefont{A.~S.} \bibnamefont{Mikhailov}},

784:   \bibinfo{author}{\bibfnamefont{V.~A.} \bibnamefont{Davydov}},

785:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{V.~S.} \bibnamefont{Zykov}},

786:   \bibinfo{journal}{Physica D} \textbf{\bibinfo{volume}{70}},

787:   \bibinfo{pages}{1} (\bibinfo{year}{1994});

788: \bibinfo{author}{\bibfnamefont{V.~A.} \bibnamefont{Davydov}},

789:   \bibinfo{author}{\bibfnamefont{V.~S.} \bibnamefont{Zykov}},

790:   \bibinfo{author}{\bibfnamefont{A.~S.} \bibnamefont{Mikhailov}},

791:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{P.~K.}

792:   \bibnamefont{Brazhnik}}, \bibinfo{journal}{Radiophys. Quantum Electron.}

793:   \textbf{\bibinfo{volume}{31}}, \bibinfo{pages}{574} (\bibinfo{year}{1988}).

794: 

795: \bibitem[{\citenamefont{Rousseau and Kapral}(2000)}]{rousseau00}

796: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Rousseau}} \bibnamefont{and}

797:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Kapral}},

798:   \bibinfo{journal}{Chaos} \textbf{\bibinfo{volume}{10}}, \bibinfo{pages}{812}

799:   (\bibinfo{year}{2000}).

800: 

801: \bibitem[{\citenamefont{Chavez et~al.}(2001)\citenamefont{Chavez, Kapral,

802:   Rousseau, and Glass}}]{chavez01}

803: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Chavez}},

804:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Kapral}},

805:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Rousseau}}, \bibnamefont{and}

806:   \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Glass}},

807:   \bibinfo{journal}{Chaos} \textbf{\bibinfo{volume}{11}}, \bibinfo{pages}{757}

808:   (\bibinfo{year}{2001}).

809: 

810: 

811: \end{thebibliography}

812: 

813: 

814: \end{document}

815: %

816: % ****** End of file template.aps ******

817: