Linear Models in Medical Imaging. John Kornak MI square February 21, 2012

Size: px

Start display at page:

Download "Linear Models in Medical Imaging. John Kornak MI square February 21, 2012"

Elwin Arnold
6 years ago
Views:

1 Linear Models in Medical Imaging John Kornak MI square February 21, 2012

2 Acknowledgement / Disclaimer Many of the slides in this lecture have been adapted from slides available in talks available on the SPM web site.

3 Overview Motivation Linear model formulation Region of interest analyses Pixel/voxel based analyses Multiple comparisons for images Bayesian image analysis methods

4 Motivation Imaging data statistical methods to look for regional effects Tissue differences between groups or over time VBM, TBM (voxel/tensor-based morphometry) PET (positron emmission tomography), fmri (functional MRI) determine activation in the brain due to thought, stimulus or task Diffusion (DWI, DTI, tractography), Bone mineral density etc. etc.

5 FMRI Data: Set of Volumes (over time) or Set of Time-Series (over space) Serial Snapshots of Volunteers brain Time Time Active Passive Baseline

6 Software etc. SPM PET, fmri, VBM and TBM, EEG/MEG ( needs Matlab) FSL fmri primarily + DTI ( R AnalyzeFMRI package + linear models in general ( and then go to your nearest CRAN mirror) Also, check Venables and Ripley Splus book + many R books (see R web site) + online tutorials

7 Challenges Generating suitable (statistical) imaging models Dealing with highly multivariate responses (curse of dimensionality) Defining imaging hypotheses Creating computationally efficient analysis procedures

8 Aims of Statistical Modeling Summarize data Estimation: point and interval estimates Inference: hypotheses / relationships Prediction

9 Aims of Statistical Modeling Summarize data Estimation: point and interval estimates Inference: hypotheses / relationships Prediction

10 Statistical Modeling Strategy Propose a model for the data Fit the model Assess the model s adequacy Fit other plausible models Compare all fitted models Interpret the best model

11 Statistical Models: Definitions Univariate response variable y i (for exp. unit i) Covariates (x i1, x i2,..., x ik ) = (variables of interest and nuisance variables) x i T Data is: { y i,x T i ;i = 1,...,n}, n experimental units Continuous covariates: e.g. age, blood pressure etc., (random or controlled) Factors: e.g. diagnosis, gender, drinking level (low, medium, high) etc.

12 The (General) Linear Model A simple linear model might take the form: e.g. y i = β 1 + x i2 β 2 + x i3 β x im β m + ε i i.i.d. = independently and identically distributed

13 The (General) Linear Model For univariate data: y i = x i T β + ε i, i = 1,...,n β = (β 1,...,β m ) T or in matrix notation y = X T β + ε is a set of unknown parameters This can be extended to a multivariate response Y = X T B + E

14 Ex. Hippocampal Volume HCV ~ Age + Diagnosis (Wilkinson notation) Diagnosis can be normal control (NC) or Alzheimer s disease (AD)

15 Ex. Hippocampal Volume HCV ~ Age + Diagnosis + Age*Diagnosis (Wilkinson notation) Diagnosis can be normal control (NC) or Alzheimer s disease (AD)

16 Structural T1 weighted MRI s Hippocampal volumes manually traced Volume measure = response for each subject Disease status encoded 1 for AD and 0 for NC (the term)

17 y i = β 1 + x i,age β age + x i,diag. β diag. + x i.age x i,diag. β inter + ε i Case 1 HCV age

18 y i = β 1 + x i,age β age + x i,diag. β diag. + x i.age x i,diag. β inter + ε i Case 2 HCV age

19 Case 3 HCV NC AD age

20 Case 4 HCV NC AD age

21 y i = β 1 + x i,age β age + x i,diag. β diag. + x i.age x i,diag. β inter + ε i Case 4 HCV NC AD age

22 Linear models can be more general - only needs to be linear in the parameters: We can have: y = x β + x 2 β + exp(x )β + x π x β + ε i age 1 age 2 height 3 age height 4 i But not y i = x age exp(β 1 + x height β 2 ) i = 1,...,n i = 1,...,n

23 Estimation Minimize squared error (Least Squares Error) = Maximum Likelihood Estimation for linear model ˆβ = (X T X) 1 X T y E( ˆβ) = β V ( ˆβ) = σ 2 (X T X) 1 Estimate by sum of squares error ˆσ 2 = n or divide by n-1 for unbiased estimate

24 Inference Model Comparison Take linear model y = X T β + ε And add constraint this defines a new model that is a simplification of the previous one

25 Inference Model Comparison E.g., cf. model y i = β 1 + β 2 x i1 + β 3 x i2 + ε i to simplification with β 3 = 0 i.e. y i = β 1 + β 2 x i + ε i β 1 (0,0,1) β 2 β 3 = 0 i.e. Aβ = c

26 What about &? β 2 = 0 β 3 = 0 Aβ = c β 1 β 2 β 3 = 0 0

27 And what about? β 2 = β 3 β 1 ( ) Aβ = c β 2 β 3 = 0 Are 2 different conditions equivalent? E.g. is the activation effect: reading a word vs imagining the object equal?

28 Definition: Linear model nested in another if 1 st model can be obtained by linear constraint on the 2 nd Nesting tree: age + gender age gender null

29 F-test for General Linear Hypothesis ( ) y = X T β + ε ε N n 0,σ 2 I n Consider This is the General Linear Hypothesis

30 Under, i.e., Aβ = c F = (SSE nested SSE larger ) / ( p larger p nested ) (SSE larger ) / (n p larger ) F plarger p nested,n p larger p denotes the number of model parameters n denotes the number of data points SSE = Deviance = sum of squares of residuals Tests whether ratio of variances = 1

31 FMRI Data: Set of Volumes (over time) or Set of Time-Series (over space) Serial Snapshots of Volunteers brain Time Time Active Passive Baseline

32 image data design matrix kernel parameter estimates realignment & motion correction smoothing Linear Model model fitting statistic image Thresholding & Random Field Theory normalisation anatomical reference Statistical Parametric Map (test statistics) Corrected thresholds & p-values

33 Estimation The estimation entails finding the parameter values such that the linear combination best fits the data Active Passive Baseline β 1 + β 2 + β 3

34 Parameter Estimates Same model for all voxels Different parameters for each voxel beta_0001.img Time-series ˆβ = ˆβ = beta_0002.img... ˆβ = beta_0003.img

35 y X T β SPM View β 1 β 2 β 1 +β 2 +β 3 β 1 +β 2 +β 3 β 3 Note: We trust: Long series with large effects and small error

36 Spatial Modeling

37 Spatial Hypotheses Question - how do we extend from standard univariate hypotheses to answering spatially motivated questions? Not easy - curse of dimensionality (millions of voxels) e.g. in 1D it makes sense to infer A is less than B, but what is the equivalent in 2D?

38 Spatial Testing Solutions Summarize the image into one dimensional quantities for testing (e.g. region of interest analysis) Consider the overall test as a combination of individual voxel tests (voxel based analysis) Perform shape/object analysis on objects defined via landmarks Build Bayesian image analysis models

39 Spatial Testing Solutions Summarize the image into one dimensional quantities for testing (e.g. region of interest analysis) Consider the overall test as a combination of individual voxel tests (voxel based analysis) Perform shape/object analysis on objects defined via landmarks Build Bayesian image analysis models

40 Voxel based analysis Each voxel obtains a test statistic from the linear model, e.g. t or F Forms statistical maps of the statistics

41 image data design matrix kernel parameter estimates realignment & motion correction smoothing Linear Model model fitting statistic image Thresholding & Random Field Theory normalisation anatomical reference Statistical Parametric Map (test statistics) Corrected thresholds & p-values

42 Hypothesis Testing Null Hypothesis H 0 Test statistic T t observed realization of T α-level Acceptable false positive risk Level α = Pr( T>u α H 0 ) Threshold u α controls false positive risk at level α u α Null Distribution of T α

43 Multiple Comparisons Problem Which of 100,000 voxels are significant? α =0.05 5,000 false positive voxels

High Threshold Med. Threshold Low Threshold t > 5.5 t > 3.5 t > 0.

44 How can we determine a sensible threshold level? Assessing Statistic Images Where s the signal or change? High Threshold Med. Threshold Low Threshold t > 5.5 t > 3.5 t > 0.5 Good Specificity Poor Power (risk of false negatives) Poor Specificity (risk of false positives) Good Power

45 Multiple Comparison Solutions: Measuring False Positives Familywise Error Rate (FWER) Familywise Error Existence of one or more false positives False Discovery Rate (FDR) FDR = E[FP/(TP+FP)] TP+FP voxels declared active, FP falsely so Realized false discovery rate: FP/(TP+FP)

46 Bonferroni Correction FWE, α, for N independent voxels is approximately α = Nv (v = voxel-wise error rate) To control FWE set v = α / N Independent Voxels Spatially Correlated Voxels Bonferroni is too conservative for brain images

FWER MCP Solutions: Random Field Theory Euler Characteristic χ u Topological Measure #blobs - #holes At high thresholds, just counts blobs Random Field

47 FWER MCP Solutions: Random Field Theory Euler Characteristic χ u Topological Measure #blobs - #holes At high thresholds, just counts blobs Random Field Threshold FWER = Pr(Max voxel u H o ) No holes = Pr(One or more blobs H o ) Never more than 1 blob Pr(χ u 1 H o ) E(χ u H o ) See description at Suprathreshold Sets

voxel size Smoothness estimation Estimate is biased when images not

48 Random Field Theory Limitations Multivariate normality (Gaussianity) Virtually impossible to check Sufficient smoothness FWHM smoothness 3-4 voxel size Smoothness estimation Estimate is biased when images not sufficiently smooth Several layers of approximations Lattice Image Data Continuous Random Field

49 Multiple Comparison Solutions: Measuring False Positives Familywise Error Rate (FWER) Familywise Error Existence of one or more false positives False Discovery Rate (FDR) FDR = E[FP/(TP+FP)] TP+FP voxels declared active, FP falsely so Realized false discovery rate: FP/(TP+FP)

50 False Discovery Rate For any threshold, all voxels can be crossclassified: Accept Null Reject Null Null True TN FP Null False FN TP N A N R Realized FDR rfdr = FP /(TP+FP) = FP /N R Special case: if N R = 0, rfdr = 0 But only can observe N R, don t know TP & FP We therefore control the expected rfdr

51 False Discovery Rate Illustration: Noise Signal Signal+Noise

52 Control of Per Comparison Rate at 10% 11.3% 11.3% 12.5% 10.8% 11.5% 10.0% 10.7% 11.2% 10.2% 9.5% Control of Familywise Error Rate at 10% Occurrence of Familywise Error FWE Control of False Discovery Rate at 10% 6.7% 10.4% 14.9% 9.3% 16.2% 13.8% 14.0% 10.5% 12.2% 8.7% Percentage of Observed Above Threshold Pixels that are False Positives

53 Benjamini & Hochberg Procedure Select desired limit q on FDR Journal of the Royal Statistical Society Series B (1995) 57: Order p-values, p (1) p (2)... p (V) Let r be largest i such that p (i) i/v q/c(v) Reject all hypotheses corresponding to p (1),..., p (r) p NB, no spatial consideration i/v p (i) i/v q/c(v)

54 Also, Non-Parametric Testing If H 0 is true then time order irrelevant (if noise really white) Therefore permute the timepoints and obtain test statistics If true test statistic is extreme compared to others then reject H 0

55 Types of Spatial Inference Individual voxel level Cluster level Set level Bayesian model based

56 Voxel-level Inference Retain voxels above α-level threshold u α Gives best spatial specificity H 0 at a single voxel can be rejected u α space Significant Voxels No significant Voxels

57 Cluster-level Inference Two step-process Define clusters by arbitrary threshold u clus Retain clusters larger than α-level threshold k α u space Cluster not significant k α k α Cluster significant

58 Cluster-level Inference Typically better sensitivity Worse spatial specificity The null hyp. of entire cluster is rejected Only means that one or more of voxels in cluster active u space Cluster not significant k α k α Cluster significant

59 Set-level Inference Count number of blobs c uses minimum blob size k to count significant activity if number of blobs > n(k,u ) Worst spatial specificity u k k space Here c = 1; only 1 cluster larger than k

60 A flexible Bayesian Approach Model the form of activity Provides an adaptive thresholding approach space Active voxels

61 Bayesian Model y = data, parameter estimates of statistics z = binary activation map modeled as a MRF x = activation level field modeled as a MRF = residual error MRF = Markov Random Field (similar random field but defined on a lattice)

62 x Model Illustration zx + z = 0 z = 1 z = 0

63 Model Illustration x zx fit z = 0 z = 1 z = 0

64 y x z zx + ε

65 Other Topics and Omissions Hemodynamic response function Multiple subjects (random and mixed effects models) PCA, ICA Multivariate analysis with variogram modeling Space-time modeling

66 Plug for: February 29 th SFASA Seminar Speaker: Time: Yoav Benjamini, PhD., Professor of Statistics Wednesday, 5pm - 6pm (4:30-5pm pre-seminar social) Location: UCSF China Basin Landing, Room 6702 (specific directions to classrooms: Title: Hierarchical Testing of Families of Hypotheses Abstract: As the size of large testing problems encountered in genomic research keeps increasing, more of these problems have further structure where the set of hypotheses can be partitioned into families of the hypotheses, and the true state of the tested signals tends to be more similar within these subsets than across the subsets. Moreover, interest may lie with a discovery of a family with some signal in it, on top of the discovery of a signal in each of the many hypotheses on its own. The challenges in the analysis of such multiple testing problems will be discussed. We then present the concept of the control on the average over the selected families of a desired error-rate, be it the family-wise error rate, the False Discovery Rate, or their generalizations. We discuss the various considerations involved using the genomic part of a Norwegian epidemiological study of breast cancer, and a study involving genomics and brain imaging. Transportation/Parking: 1) Right at the Caltrain station. 2) If driving, 2-hour parking in the neighborhood as well as reasonably priced parking in lot A for the Giants. The building parking is expensive. 3) If by BART, transfer at Embarcadero to the N-Judah or T-train (both inbound) 66

Linear Models in Medical Imaging. John Kornak MI square February 22, 2011

Linear Models in Medical Imaging John Kornak MI square February 22, 2011 Acknowledgement / Disclaimer Many of the slides in this lecture have been adapted from slides available in talks available on the