Random shapes in brain mapping and astrophysics using an idea from geostatistics Keith Worsley, McGill Jonathan Taylor, Stanford and Université de Montréal Arnaud Charil, Montreal Neurological Institute CfA red shift survey, FWHM=13.3 100 80 Euler Characteristic (EC) 60 "Meat ball" topology 40 20 "Bubble" topology 0 -20 -40 "Sponge" topology -60 -80 -100 -5 CfA Random Expected -4 -3 -2 -1 0 1 Gaussian threshold 2 3 4 5 Brain imaging Detect sparse regions of “activation” Construct a test statistic image for detecting activation Activated regions: test statistic > threshold Choose threshold to control false positive rate to say 0.05 i.e. P(max test statistic > threshold) = 0.05 Bonferroni???  ¹= max Z1 cos µ + Z2 sin µ 0·µ·¼=2 Example test statistic: Z1~N(0,1) Z2~N(0,1) s2 3 2 1 0 -1 -2 Excursion sets, Xt = fs :  ¹ ¸ tg s1 -3 Threshold t 4 Rejection regions, Z2 2 Search Region, S Rt = fZ :  ¹ ¸ tg 3 2 1 0 Z1 -2 -2 0 2 Euler characteristic heuristic Search Region, S Euler characteristic, EC EC= 1 7 Excursion sets, Xt 6 5 2 1 1 0 10 Observed 8 ¸ t) P(max Â(s) ¹ s2S 6 ¼ E(EC) = 0:05 Expected ) t = 3:75 4 2 0 -2 0 0.5 1 1.5 E(EC(S \ Xt )) = 2 X D d=0 2.5 L (S)½ (t) d d 3 3.5 4 Threshold, t E(EC(S \ Xt )) = X D d=0 Tube(λS,r) Radius, r Tube(Rt,r) 14 L (S)½ (t) d d r 12 10 µ @Z @s ¸ = Sd p 4 log 2 = FWHM ¶ r 0 1 4 0.5 2 Z1 0.2 Area 0 jTube(¸S; r)j = X D d=0 2L1 (S)r ¼ L0 (S)r2 0 0.5 1 1.5 2 Radius of Tube, r ¼d L ¡d (S)r d D ¡(d=2 + 1) Probability L (S) 2 50 0 EC ½d (t) -2 density 0 2 0.4 jTube(¸S; r)j 100 0.4 -2 L (S)0 Lipschitz-Killing 2 4 6 curvature 8 10 12 14 d 150 0.8 0.6 λS 6 Rt 2 1.5 8 1 Z2 2 Radius, r P(Tube(Rt ; r)) 0.3 p 2¼½1 (t)r ½0 (t) ¼½2 (t)r2 0.2 0.1 0 0 0.5 Radius of 1 Tube, r 1 X P(Tube(Rt ; r)) = (2¼)d=2 ½d (t)rd =d! d=0 EC density ½d (t) of the  ¹ statistic Z2~N(0,1) Tube(Rt,r) r Rejection region Rt t-r t Z1~N(0,1) Taylor’s Gaussian Kinematic Formula: 1 X P (Z1 ; Z2 2 Tube(Rt ; r)) = (2¼)d=2 ½d (t)rd =d! d=0 ½0 (t) = Z = ½0 (t) + (2¼)1=2 ½1 (t)r + (2¼)½2 (t)r2 =2 + ¢ ¢ ¢ Z 1 = (2¼)¡1=2 e¡z2 =2 dz + e¡(t¡r)2 =2 =4 t¡r 1 (2¼)¡1=2 e¡z2 =2 dz + e¡t2 =2 =4 t ½1 (t) = (2¼)¡1 e¡t2 =2 + (2¼)¡1=2 e¡t2 =2 t=4 ½2 (t) = (2¼)¡3=2 e¡t2 =2 t + (2¼)¡1 e¡t2 =2 (t2 ¡ 1)=8 .. . Lipschitz-Killing curvature Ld (S) r Tube(λS,r) λS Steiner-Weyl Volume of Tubes Formula: Area(Tube(¸S; r)) = X D ¼ d=2 L ¡d (S)r d D ¡(d=2 + 1) d=0 = L2 (S) + 2L1 (S)r + ¼ L0 (S)r2 = Area(¸S) + Perimeter(¸S)r + EC(¸S)¼r2 L (S) = EC(¸S) = Resels0 (S) 0 p L (S) = 1 Perimeter(¸S) = 4 log 2 Resels (S) 1 1 2 L (S) = Area(¸S) = 4 log 2 Resels2 (S) 2 How to ¯nd Lipschitz-Killing curvature Ld (S) Edge length × λ FWHM/√(4log2) 12 10 8 6 4 2 . .. . . . . . . .. . . .. . . . . . . . . . . . . . . . .. . . . . . ... . . 4 .. . . . . . . . . . . . 6 .. . . . . . . . . . . . . . . . . . . . . . . . 8 .. . . . ... . .. . . . . . . . ..... . . . . .... .. .. 10 of simplices L (Lipschitz-Killing ²) = 1, L (¡) curvature L (N = 1, )=1 0 0 0 L (¡) = edge length, L (N) = 1 perimeter 1 2 L1 (N) = area 2 P Lcurvature P L Lipschitz-Killing union L ² ¡ Pof L ¡ of simplices N (S) = P² 0 ( ) ¡ 0( ) + P L (S) = L (¡) ¡ L (N) ¡ N 1 L1 (S) = P L 1(N) 2 N 2 0 N 0 ( ) Non-isotropic data? µ ¸ = Sd Z~N(0,1) s2 3 @Z @s ¶ p = 4 log 2 FWHM 2 1 0.14 0.12 0 -1 -2 s1 12 10 8 6 4 2 . ... . . . . . .. . . .. . . . . . . . . . . . . . . . .. . . . . . ... . . 4 .. . . . . . . . . . . . 6 -3 Can we warp the data to isotropy? .. . i.e. multiply edge lengths by λ? . . .. . . . . . . . ... Locally . . . . . yes, but we may need extra dimensions. . . . . . . . . . . . . . ..... Nash Embedding Theorem: . . . . . dimensions ≤ D + D(D+1)/2 . . . .... . . ... . . . D=2: dimensions ≤ 5 8 10 0.1 0.08 0.06 Warping to isotropy not needed – only warp the ptriangles µ ¶ ¸ = Sd Z~N(0,1) s2 3 @Z @s = 4 log 2 FWHM 2 1 0.14 0.12 0 -1 -2 Edge length × λ FWHM/√(4log2) 12 10 8 6 4 2 .. . . . . . . . . .. . . .. . . . . . . . . . . . . . . . .. . . . . . ... . . . . . . . . . . . . . . 4 6 .. . . . . . . . . . . . . . . . . . . . . . . . 8 . .. . . . ... . .. . . . . . . . ..... . . . . .... ... 10 0.1 0.08 0.06 -3 s1 of simplices L (Lipschitz-Killing ²) = 1, L (¡) curvature L (N = 1, )=1 0 0 0 L (¡) = edge length, L (N) = 1 perimeter 1 2 L1 (N) = area 2 P Lcurvature P L Lipschitz-Killing union L ² ¡ Pof L ¡ of simplices N (S) = P² 0 ( ) ¡ 0( ) + P L (S) = L (¡) ¡ L (N) ¡ N 1 L1 (S) = P L 1(N) 2 N 2 0 N 0 ( ) Estimating Lipschitz-Killing curvature Ld (S) We need independent & identically distributed random fields e.g. residuals from a linear model Z1 Z2 Z3 Z4 Z5 Z6 Z7 Z9 … Zn Z8 Replace coordinates of the simplices in S⊂RealD by (Z1,…,Zn) / ||(Z1,…,Zn)|| in Realn of simplices L (Lipschitz-Killing ²) = 1, L (¡) curvature L (N = 1, )=1 0 0 0 L (¡) = edge length, L (N) = 1 perimeter 1 2 L1 (N) = area 2 P Lcurvature P L Lipschitz-Killing union L ² ¡ Pof L ¡ of simplices N (S) = P² 0 ( ) ¡ 0( ) + P L (S) = L (¡) ¡ L (N) ¡ N 1 L1 (S) = P L 1(N) 2 N 2 0 N 0 Unbiased! ( ) Unbiased! MS lesions and cortical thickness Idea: MS lesions interrupt neuronal signals, causing thinning in down-stream cortex Data: n = 425 mild MS patients Lesion density, smoothed 10mm Cortical thickness, smoothed 20mm Find connectivity i.e. find voxels in 3D, nodes in 2D with high correlation(lesion density, cortical thickness) Look for high negative correlations … n=425 subjects, correlation = -0.568 Average cortical thickness 5.5 5 4.5 4 3.5 3 2.5 2 1.5 0 10 20 30 40 50 60 Average lesion volume 70 80 Thresholding? Cross correlation random field Correlation between 2 fields at 2 different locations, searched over all pairs of locations one in R (D dimensions), one in S (E dimensions) sample size n Cao & Worsley, Annals of Applied Probability (1999) MS lesion data: P=0.05, c=0.325, T=7.07 Normalization LD=lesion density, CT=cortical thickness Simple correlation: Subtracting global mean thickness: Cor( LD, CT ) Cor( LD, CT – avsurf(CT) ) And removing overall lesion effect: Cor( LD – avWM(LD), CT – avsurf(CT) ) 0.1 correlation 0 5 x 10 2.5 2 -0.1 1.5 -0.2 1 -0.3 threshold -0.4 -0.5 0 50 100 150 Different hemisphere 0.1 5 x 10 2.5 0 correlation Histogram Same hemisphere -0.1 2 -0.2 1.5 -0.3 1 0.5 -0.4 0 -0.5 0 threshold 50 100 150 0.5 0 ‘Conditional’ histogram: scaled to same max at each distance 0.1 1 -0.1 0.6 -0.2 0.4 -0.3 -0.4 -0.5 0 threshold 50 100 150 distance (mm) 1 0 0.8 correlation correlation 0 0.1 0.8 -0.1 0.6 -0.2 0.4 -0.3 0.2 -0.4 0 -0.5 0 threshold 50 100 150 distance (mm) 0.2 0