Fred L. Bookstein

Principal warps: thin-plate splines and the decomposition of deformations

The decomposition of deformations by principal warps is demonstrated. The method is extended to deal with curving edges between landmarks. This formulation is related to other applications of splines current in computer vision. How they might aid in the extraction of features for analysis, comparison, and diagnosis of biological and medical images in indicated. >

Landmark methods for forms without landmarks: morphometrics of group differences in outline shape

Morphometrics, a new branch of statistics, combines tools from geometry, computer graphics and biometrics in techniques for the multivariate analysis of biological shape variation. Although medical image analysts typically prefer to represent scenes by way of curving outlines or surfaces, the most recent developments in this associated statistical methodology have emphasized the domain of landmark data: size and shape of configurations of discrete, named points in two or three dimensions. This paper introduces a combination of Procrustes analysis and thin-plate splines, the two most powerful tools of landmark-based morphometrics, for multivariate analysis of curving outlines in samples of biomedical images. The thin-plate spline is used to assign point-to-point correspondences, called semi-landmarks, between curves of similar but variable shape, while the standard algorithm for Procrustes shape averages and shape coordinates is altered to accord with the ways in which semi-landmarks formally differ from more traditional landmark loci. Subsequent multivariate statistics and visualization proceed mainly as in the landmark-based methods. The combination provides a range of complementary filters, from high pass to low pass, for effects on outline shape in grouped studies. The low-pass version is based on the spectrum of the spline, the high pass, on a familiar special case of Procrustes analysis. This hybrid method is demonstrated in a comparison of the shape of the corpus callosum from mid-sagittal sections of MRI of 25 human brains, 12 normal and 13 with schizophrenia.

Spatial Pattern Analysis of Functional Brain Images Using Partial Least Squares

This paper introduces a new tool for functional neuroimage analysis: partial least squares (PLS). It is unique as a multivariate method in its choice of emphasis for analysis, that being the covariance between brain images and exogenous blocks representing either the experiment design or some behavioral measure. What emerges are spatial patterns of brain activity that represent the optimal association between the images and either of the blocks. This process differs substantially from other multivariate methods in that rather than attempting to predict the individual values of the image pixels, PLS attempts to explain the relation between image pixels and task or behavior. Data from a face encoding and recognition PET rCBF study are used to illustrate two types of PLS analysis: an activation analysis of task with images and a brain-behavior analysis. The commonalities across the two analyses are suggestive of a general face memory network differentially engaged during encoding and recognition. PLS thus serves as an important extension by extracting new information from imaging data that is not accessible through other currently used univariate and multivariate image analysis tools.

Incidence of fetal alcohol syndrome and prevalence of alcohol-related neurodevelopmental disorder.

We critique published incidences for fetal alcohol syndrome (FAS) and present new estimates of the incidence of FAS and the prevalence of alcohol-related neurodevelopmental disorder (ARND). We first review criteria necessary for valid estimation of FAS incidence. Estimates for three population-based studies that best meet these criteria are reported with adjustment for underascertainment of highly exposed cases. As a result, in 1975 in Seattle, the incidence of FAS can be estimated as at least 2.8/1000 live births, and for 1979-81 in Cleveland, approximately 4.6/1,000. In Roubaix, France (for data covering periods from 1977-1990), the rate is between 1.3 and 4.8/1,000, depending on the severity of effects used as diagnostic criteria. Utilizing the longitudinal neurobehavioral database of the Seattle study, we propose an operationalization of the Institute of Medicines recent definition of ARND and estimate its prevalence in Seattle for the period 1975-1981. The combined rate of FAS and ARND is thus estimated to be at least 9.1/1,000. This conservative rate--nearly one in every 100 live births--confirms the perception of many health professionals that fetal alcohol exposure is a serious problem.

Risk factors for adverse life outcomes in fetal alcohol syndrome and fetal alcohol effects.

ABSTRACT. Clinical descriptions of patients with Fetal Alcohol Syndrome (FAS) and Fetal Alcohol Effects (FAE) suggest major problems with adaptive behavior. Five operationally defined adverse outcomes and 18 associated risk/protective factors were examined using a Life History Interview with knowledgeable informants of 415 patients with FAS or FAE (median age 14 years, range 6-51; median IQ 86, range 29-126). Eighty percent of these patients were not raised by their biological mothers. For adolescents and adults, the life span prevalence was 61% for Disrupted School Experiences, 60% for Trouble with the Law, 50% for Confinement (in detention, jail, prison, or a psychiatric or alcohol/drug inpatient setting), 49% for Inappropriate Sexual Behaviors on repeated occasions, and 35% for Alcohol/Drug Problems. The odds of escaping these adverse life outcomes are increased 2- to 4-fold by receiving the diagnosis of FAS or FAE at an earlier age and by being reared in good stable environments.

Fitting conic sections to scattered data

The problem of fitting conic sections to scattered data has arisen in several applied literatures. The quadratic from Ax 2 + Bxy + Cy 2 + Dx + Ey + F that is minimized in mean-square is proportional to the ratio of two squared distances along rays through the center of a conic. Considerations of invariance under translation, rotation, and scaling of the data configuration lead to a straightforward method of estimation somewhat different from earlier suggestions. The method permits an extension to conic splines around extended digitized curves, expediting a smooth reconstruction of their curvature. Some examples are presented indicating how the technique might be applied in morphometrics.

Preserved CD4+ Central Memory T Cells and Survival in Vaccinated SIV-Challenged Monkeys

Vaccine-induced cellular immunity controls virus replication in simian immunodeficiency virus (SIV)–infected monkeys only transiently, leading to the question of whether such vaccines for AIDS will be effective. We immunized monkeys with plasmid DNA and replication-defective adenoviral vectors encoding SIV proteins and then challenged them with pathogenic SIV. Although these monkeys demonstrated a reduction in viremia restricted to the early phase of SIV infection, they showed a prolonged survival. This survival was associated with preserved central memory CD4+ T lymphocytes and could be predicted by the magnitude of the vaccine-induced cellular immune response. These immune correlates of vaccine efficacy should guide the evaluation of AIDS vaccines in humans.

Combining the Tools of Geometric Morphometrics

The greatest strength of the new geometric morphometrics is the system of interrelated multivariate and graphical procedures it offers for a variety of analytic questions involving landmark data. A typical analysis will begin with the conversion of landmark data into a multivariate statistical representation of shape, will continue with a series of broadly familiar multivariate matrix manipulations, and will conclude by inspection of a considerable variety of diagrams that represent the findings in both the space of shape coordinates per se and the space of the two-or three-dimensional image of the organism. The choices under the first heading, the passage to a multivariate representation of shape, include two-point shape coordinates, partial warp scores, and Procrustes residuals. Each of these except the partial warp scores is unsuitable for some subset of the reasonable matrix manipulations; for instance, shape coordinates do not supply sensible principal components analyses, and Procrustes residuals cannot lead to sound canonical variate analyses without modification. The modes of diagramming data include thin-plate splines, partial warp splines and scatters, Procrustes residual scatters, and resistant-fit scatters, among others. Most analyses benefit greatly from exploiting more than one of these.

Cranial integration in Homo: singular warps analysis of the midsagittal plane in ontogeny and evolution

This study addresses some enduring issues of ontogenetic and evolutionary integration in the form of the hominid cranium. Our sample consists of 38 crania: 20 modern adult Homo sapiens, 14 sub-adult H. sapiens, and four archaic Homo. All specimens were CT-scanned except for two infant H. sapiens, who were imaged by MR instead. For each specimen 84 landmarks and semi-landmarks were located on the midsagittal plane and converted to Procrustes shape coordinates. Integration was quantified by the method of singular warps, a new geometric-statistical approach to visualizing correlations among regions. The two classic patterns of integration, evolutionary and ontogenetic, were jointly explored by comparing analyses of overlapping subsamples that span ranges of different hypothetical factors. Evolutionary integration is expressed in the subsample of 24 adult Homo, and ontogenetic integration in the subsample of 34 H. sapiens. In this data set, vault, cranial base, and face show striking and localized patterns of covariation over ontogeny, similar but not identical to the patterns seen over evolution. The principal differences between ontogeny and phylogeny pertain to the cranial base. There is also a component of cranial length to height ratio not reducible to either process. Our methodology allows a separation of these independent processes (and their impact on cranial shape) that conventional methods have not found.

Comparing frontal cranial profiles in archaic and modern homo by morphometric analysis

Archaic and modern human frontal bones are known to be quite distinct externally, by both conventional visual and metric evaluation. Internally this area of the skull has been considerably less well‐studied. Here we present results from a comparison of interior, as well as exterior, frontal bone profiles from CT scans of five mid‐Pleistocene and Neanderthal crania and 16 modern humans. Analysis was by a new morphometric method, Procrustes analysis of semi‐landmarks, that permits the statistical comparison of curves between landmarks. As expected, we found substantial external differences between archaic and modern samples, differences that are mainly confined to the region around the brow ridge. However, in the inner median‐sagittal profile, the shape remained remarkably stable over all 21 specimens. This implies that no significant alteration in this region has taken place over a period of a half‐million years or more of evolution, even as considerable external change occurred within the hominid clade spanning several species. This confirms that the forms of the inner and outer aspects of the human frontal bone are determined by entirely independent factors, and further indicates unexpected stability in anterior brain morphology over the period during which modern human cognitive capacities emerged. Anat Rec (New Anat): 257:217–224, 1999.