First we must relate the misadventures of the mummy of Ramses II from the death of the pharaoh until his arrival at Chaillot Palace.

Biostereometrical approach in aid of moire interference fringe has an advantage in that the contour fringes on the surface of an object are promptly visible to the naked eye and are recorded photographically. Even the movement of these fringes concomitant with the motion of the human body can be recorded by movie camera or video TV. One of the limitations of this moire method using radiating light is that the value of contour interval is obliged to increase gradually with the advancement of the ordinal number of fringes counted from the grating. To cover this demerit, a new moire technique with parallel light has been developed at our laboratory. The parallel light is provided by an optical system including a field lense. Our optical apparatus makes it possible to get equal intervals of the succeeding contour fringes, irrespective of the fringe order. Any requested contour intervals are attained by choosing a grating with suitable pitch and/or an appropriate incidence angle of the parallel light at the grating. The automatic input of analogue data of moire patterns into the central processing unit is accomplished by a TV camera and A/D converter. Examples of line printer output through this conversion system are given.

The display stereocomparator, which was primarily designed for the evaluation of stereoradiographs, is quite similar to the common stereocomparator, except that the usual point-shaped target is replaced by a complete stereoscopic image pair of a model body. The model body is superimposed to the spatial image by semi-transparent mirrors. The image pair of the model body is calculated by a computer simulation of the stereoscopic projection and is displayed on cathode ray tubes which are on-line connected to the computer. Also by computer simulation the model body can be shifted and rotated in three dimensions, and its shape and size can be altered. Motions and deformations can be continuously controlled by interactive data input with potentiometers. Thus the model can be fitted to the structures seen in the spatial image; this is equivalent to a three-dimensional measurement. The advantages of the display-stereocomparator, especially in the case of irregular bio-logic structures are discussed. Applications of the device are demonstrated for the measurement of scoliotic spines and for the fitting of implants. Furthermore the display-stereocomparator can be used in some non-stereoscopic measurements. An example of this application using monocular vision of the two images is given in the paper of B. Drerup and W. Frobin (Munster).

The fundamental characteristics of analytical stereoplotters compared to the analog stereoplotters is their facility to use them and the versatility given by the on-line of a mini-computer. Analog stereoplotters, even those classified universal, are generally designed at the beginning for cartographic plotting from aerial photographies. Few of them have the capability to be used in non cartographic field such as biostereometrics applications. Analytical stereoplotters are usable in all kind of restitution, monoculary or stereoscopically. The main particularity of TRASTER, the analytical stereoplotter built by Matra, is its stereo viewing system obtained by an optical projection in polarized light of the photographies in place of the conventionnal binocular system. This feature increases the operator comfort, gives a better facility of the stereophoto-grammetry adaptation to not confirmed operators and allows the stereo view by several people in the same time. Thanks to the incorporated computer, specific softwares can be easily adapted to Traster. Restitution of stereopairs are possible in on-line mode on the drafting table. Three dimensional coordinates of objects to be plotted can be stored in numerical form for further processing.

A simplified stereophotogrammetric method for the measurement of body surfaces is presented. The advantage of this method in comparison with common stereophotogrammetry is that only one photograph has be evaluated. The technical equipment consists of a projector with a grid slide and a camera. This is equivalent to the common arrangement for stereophotography, if the directions of the light rays emerging from the projector are thought of having been reversed. So the grid slide and the photograph form some sort of a stereo image pair. The grid is projected on the surface to be measured. The evaluation of the photographs is accomplished by the identification of corresponding points. The correspondance can easily be established by counting the crosspoints of the grid. The coordinates of the grid points in the slide can be calculated by this count, whereas those in the photograph have to be measured. Thus a correlation of two images by stereoscopic vision is not necessary. The measurement can be performed with a simple di-gitizer.The calibration of the apparatus is done by using landmarks and solving the normal equations for the orientation elements of the camera and of the projector. With a slight modification the method can be employed for quasi-continuous measurements of the body surface along projected grid lines.

The method here described allows a quantitative study by photonic microscopy of thick material without injuring its structure. A tridimensional photographic model is built through superposing in parallel planes a set of microphotographic transparencies of successive optical cross-optical through the material. This stereomodel exactly features the spatial structure of the object and allows, either directly or by means of a stereogram, to measure its tridimensional coordinates. These data are computerized to obtain linear and angular measurements, and also to produce sets of automacic successive perspective drawings of the object, obtained by rotating the reference axes. The microscope must be fitted up with objectives with a wide numerical aperture and a plane field. An interferential device controls the focusing. The method has been initially worked out to study metaphasic chromosomes in situ within the cell without using any of the customary techniques : squash, hypotonic shock, cross-sections, etc..., which result in a definite loss of information. Basically it is applicable to most of the biological materials, and has produced, as it seems for the fist time in 1968, a tridimensional quantitative study of cellular components by photonic microscopy.

Studying the breathing movements of human trurick and abdomen during inspiration and expiration, we obtained spatiotemporal informations. With morphometric drawings, it has been possible to build sculptures which reveal an accurate scientific object as well a work of art.

Although Democritus, a Greek pholosopher of the fifth century B.C. described the use of cross-sections in analyzing a solid form, this method was not extensively developed in art until the Renaissance. The earliest treatise documenting the integration of the cross-section and linear perspective is Piero della Francesca's De prospective pingendi (c. 1480), in which a drawing of the human head is mathematically conceived and plotted by means of cross-section contours. Piero's method anticipates contemporary biostereometric techniques and current theories of visual perception. Outside of theoretical treatises the complete cross-section rarely occurs in art, though certain pictorial elements such as the religious halo can be interpreted as cross-sections. The chan-ging representation of the halo in art of the Medieval, Renaissance and Baroque periods parallels the development of the artist's concepts and techniques for representing form and space. During the Renaissance and Baroque periods the widespread use of contour hatching, a drawing technique based on the cross-section, indicates that the cross-section concept has played a greater role in pictorial representation than has generally been recognized.

A study of improved computer graphic representation of the human figure is being conducted under a National Science Foundation grant. Special emphasis is given biostereometrics as a primary data base from which applications requiring a variety of levels of detail may be prepared. For example, a human figure represented by a single point can be very useful in overview plots of a population. A crude ten point figure can be adequate for queuing theory studies and simulated movement of groups. A one hundred point figure can usefully be animated to achieve different overall body activities including male and female figures. A one thousand point figure si-milarly animated, begins to be useful in anthropometrics and kinesiology gross body movements. Extrapolations of this order-of-magnitude approach ultimately should achieve very complex data bases and a program which automatically selects the correct level of detail for the task at hand. See Summary Figure 1.

Biostereometrics transcribes a biological surface into a large sample of coordinate triples for the ordinary points borne there. For measurement of shape, the accidents of this sampling must be eliminated and the transcription replaced by a few geometric descriptors. More of these latter are available for the task of significant morphometric comparison than the literature acknowledges. For instance, the information to be gleaned from each landmark on the surface includes, in addition to the locus itself, at least eight more parameters of the curving surface there. These concisely represent bulges, ridges, sags, and other ordinary features of form by new tensor descriptors which, though wholly different-looking from the biostereometric coordinate record, derive therefrom. In this analysis, growth is distortion of one surface into another, and its natural descriptor, expressing how all loci are growing apart from each other, is the symmetrized derivative of the distortion mapping.

A brief discussion of some of the problems inherent in the quantitative description of complex morphological forms is presented. These include: 1) an accurate measurement of shape, 2) the removal of the confounding effect of size and 3) the application of the procedures of positional orientation and size standardization so that morphological forms can be meaningfully compared. The suitability of Fourier analysis to characterize complex forms of the type encountered in biology is described.

We describe some of our recent work on the computer-assisted acquisition, interaction, manipulation and display of three-dimensional data in general, and the human body form in particular. The work encompasses four distinct areas: 1) the automatic acquisition of 3-D surface points using a light-scanning technique, 2) the automatic construction of mathematical surface descriptions from point-contour data, 3) interaction and manipulation of 3-D data with multiple hand-mounted sensors, and 4) real-time stereo display of computer-synthetized continuous-tone images on special head-mounted display devices. We present some of our preliminary results in each of these areas and outline how they will be combined into a single comprehensive system. This system we expect will significantly enhance the computer's capabilities to assist in working with complex three-dimensional structures such as human bodies. As a simple example, we expect the system should be able to automatically digitize a human model, and within a few minutes generate a three-space approximating surface conforming to the human model's surface. Another person, wearing a helmet-like display device, will see through his helmet's visor not only the human model, but also the computer-generated approximating surface superimposed on it. As this human viewer moves around the model, the computer-generated surface will appear to stay on the model, and he'll be able to reach out and modify any portion of it or indicate that the digitizing system should rescan parts of the sur-face whose approximation he deems unsatisfactory. Other applications of the system for diagnostic medicine, architecture, and molecular modeling will also be suggested.

1 - Requisite conditions for acquisition of photostereoscopic data for analytic plotting: . absence of servitudes with respect to camera characteristics, . rigorous limitation of constraints imposed on subject by control points, . possibility of processing through multiple or heterogenous optical media, . possibility of achieving systems of stereopairs around a subject, . examples of conditions for space measurement of shapes in biology. 2 - Processing data on analytic plotters : . plotting by a biometrics specialist without photogrammetric training, . recording of point co-ordinates in magnetic storage, . simultaneous graphics on datum plan chosen at random, . immediate resetting of previously processed stereopairs, . use of either macro- or microscopic stereopairs, indifferently. 3 - Conclusion : . non-limitative acquisition and processing system, . simple direct use by specialized physicians, . eventual capability of adaptation to specific new problems.

Failure of total joint replacement due to loosening of the composents either between the implant and cement or between the cement and bone is emerging as a late complication with an incidence as high as 20 percent. Loosening may not only cause pain but progressive loss of support for the prosthesis with eventual structural failure. Early diagnosis is important so that revision may be carried when deterioration or pain occurs. No method is currently available which clearly establishes loosening at an early stage except surgical exploration. We have devised a method based on our in vivo photogrammetry studies of patellar tracking patterns using metallic markers placed in bone near both components of the total joint. Stereo x-rays taken with the joint loaded and unloaded are measured for relative motion between the implant and the metallic markers. Laboratory studies using prosthetic hip components mounted in plastic bone have revealed the ability of this method to detect pistoning movements as small as 80 microns. These findings were confirmed by physical measurements.

Loosening of the prosthetic device occurs in about 5% of cases following placement of total hip prostheses (THP). Early detection of loosening is much desired but is difficult to achieve using conventional methods. Due to errors of projection, it is quite possible to fail to detect mobility of even as much as 5 mm on single x-ray films. We are attempting to develop a simplified photogrammetric system suitable for general hospital use which could detect loosening of 0.8 mm at the 95 % level of confidence without use of complex stereoplotting equipment. Metal reference markers are placed in the shaft of the femur and in the acetabular region of the pelvis at the time of surgery. The distances between these reference markers and certain unambiguous points on the prostheses are computed analytically using an X-Y acoustical digitizer (accuracy ± 0.1 mm) and software developed previously for craniofacial measurement. Separate stereopairs of the joint region are taken under weight-bearing and nonweight-bearing conditions. Differences in the measured distances between the bo-ne markers and the prosthetic components on the two stereopairs are taken as indicators of prosthesis loosening. Measurements on a phantom using ten different x-ray stereopairs taken from as many different perspectives have established that true linear distances between reference points and prostheses can be measured at the desired reliability with the present low precision system. Preliminary in vivo measurements indicate that the main unresolved problem is the movement of the subject between the two exposures of each single stereopair. Two possible solutions to this problem are discussed.

The use of a flat calibration object with a known 2-D control point distribution for simultaneous estimation of internal and external camera parameters in whole-body movement studies allows coverage of the full observation area as required to avoid extrapolation errors in subsequent 3-D trajectory reconstruction, without the cumbersomeness of more conventional 3-D calibration objects designed to cover the whole observation area. In this adaptation of Simultaneous Multiframe Analytical Calibration (S.M.A.C.), the use of certain oblique (convergent) observations or frames on a planar control distribution allows full recovery of all internal camera parameters, by simultaneous estimation of the various frame positions and attitudes with respect to the camera. Calibration of a second camera whether real or virtual as in mirror photography may be conducted by observing at least one frame through both cameras. The procedure is initialized by means of the fractional linear transformation describing central projection between two planes; the only a priori information required in this approach is an approximate value for the camera principal point; no initial values are generally required for camera principal distance or for the various frame positions and attitudes. In the present paper, the theory and implementation of planar S.M.A.C. are discussed, followed by results from simulation studies. Experimental work is currently in progress.

The data acquisition system employed in our laboratory includes optical, electronic and computer subsystems. Three movie camera freeze the motion for analysis. The film is displayed on a motion analyzer, and the body segment positions are recorded in a three dimensional coordinate system with Graf/pen sonic digitizer. The angular rotations are calculated by computer and automatically plotted. The force plate provides measurements of vertical force, foreaft shear, medial-lateral shear, torque, and center of pressure. Electromyograms are superimposed upon gait movies to permit measurement of muscle phasic activity. The Hycam movie camera si-multaneously films (through separate lens) the subject and oscilloscope. Movement measurements, electromyograms, and floor reaction forces provide the data base for analysis. From a study of the gait changes in five normal subjects following tibial nerve block, and from additional studies of patients with paralysis of the ankle plantar flexors, the pathomechanics of calcaneal gait can be described. Inability to transfer weight to the forward part of the foot produces ankle instability and reduction of contralateral step length. Excessive drop of the center of mass necessitates com-pensatory increased lift energy output through the sound limb to restore the height of the center of mass. Excessive stance phase ankle dorsiflexion produces knee instability requiring prolonged quadriceps muscle phasic activity.

Detailed knowledge on normal and pathological joint motion during physiological activities such as gait is necessary both for planning and evaluating of orthopaedic surgery in pyramidal-extrapyramidal disorders of movement. A methodology of stereometric gait evaluation using cinephotography in three planes combined with a pair of transparent force plates and electro-myographic recording is described. Resultats of a foot load study in 35 normal preschool children and long them follow-up evaluations on hip and knee joint operations in cerebral spastic patients are demonstrated. Problems of picture analysis are discussed. As a mean for improvement in surgical procedures for such patients, detailed information are obtainable by cinephotography is mandatory. In addition, three-dimensional motions in large joints such as hips, knees and talo-crural articulations must be measured quantitatively with a high degree of accuracy. Special polyhedral reflective targets were developed to permit semiautomatic and automatic image analysis for joint range measurement in the lower extremities.

A two year pilot program for biostereometric analysis of treatment effectiveness in five patients with chronic decompensating back strain has been completed. The patients came from the investigators family practice of osteopathic medicine. They all manifested objective signs of ligamentous and muscular strain of their postural biomechanics due to the combined effects of prior injury to the musculoskeletal system, gravity strain and the passage of time. Two of the patients were treated with osteopathic manipulative treatment plus a pelvic leverage treatment device developed by Martin Jungman, M.D. Two patients received osteopathic treatment alone and the fifth individual switched from control to full program status in the middle of the study after the second stereophotography recording. Signs and symptoms of all patients' gravity strain syndrome changed during the program. Those patients who had the full combination of treatment modalities showed the most positive and significant postural changes as demonstrated by the biostereometric technique developed and performed by the Department of Biostereometrics, Texas Institute of Rehabilitation and Research, Baylor College of Medicine, Houston, Texas. Improvement was clearly demonstrated more quickly than with the prior radiographic measuring methods. X-ray and other studies have also been done on this group. All of the data has not been processed yet in this program. The test patients have improved posture, muscle mass and tone, more stamina and reduced pain.

To determine the activity space of a sitting subject, it is necessary to go beyond the mere statistical description of morphology and the knowledge of the displacement volume. An anlysis of the positions or variations of the positions of the diverse segmental elements (arms, hands, lower limbs, etc...) in the course of a given activity is required. Of the various methods used to locate quickly and accurately the spatial positions of anatomical points, stereometry makes it possible to plot the three-dimensional coordinates of any point in space in relation to a fixed trirectangle frame of reference determined by the stereome-tric measuring device. Thus, regardless of the orientation and posture of the subject, his segmental elements can be easily pin-pointed, throughout the experiment, within the space they occupy. Using this method, it is possible for a sample of operators seated at an operation station and applying either manual controls or pedals and belonging to a population statistically defined from the data collected and the analyses produced by the anthropometric study to determine a contour line of reach capability marking out the usable working space and to know, within this working space, a contour line of preferential activity that is limited, in space, by the whole range of optimal reach capability of all the subjects.

iostereometric measurements of 3- to 7-years old children are used to obtain equations for the mass distribution characteristics of children. Stature and weight are the independentparameters necessary for use with the equations.

The Centre Technique du Cuir (Leather Technical Center) has been entrusted with the task of measuring children's feet. A new equipement has been devised which makes the precision measures sure and which is quick to give informations. The paper will present : 1 - the existing engineerings, 2 - the research's and analysis's methodology, 3 - the CTC apparatus actually used in schools.

The points the movements of which are to be registered are marked by little light sensitive photoamplifiers. These are periodically (50 times per second) exposed to a VÃ-shaped bright figure projected by a rotating mirror. The pattern of the resulting electrical impulses is analysed by a computer controlled fast digital time counting apparatus with a clock of 100 ns. Is this way we get every 20 ms the respective position of the marked points (up to nine). The results of this method are simular to the well known photographic chronocyclography with the advantage, however, that the momentary positions of the points are calculated on line (accuracy ± 2 mm). In the intervals between the sweeps of the light-figure the computer picks up 8 lines of analogue data, actually the electromyogram of eight different muscles, and stores them, together with the optical data on digital magnetic tape for further evaluation. After the investigation the obtained data are analysed in the following way : 1) Plotting several parameters of positions (angles, c.g.) together with the electromyogram ver-sus time during one step of walking. 2) Calculating the correlation between the patterns of electrical activity of different muscles. 3) Comparison of the results of 1) and 2) between healthy persons and handicapped persons (C.P.), scoliosis, desease of spine, hip and pelvis, etc...).

X-ray photogrammetry has been applied to reconstruct the disposition of the radioactive sources in a patient to be treated by brachitheraphy for uterus cancer.

Since 1972 a system for analytic roentgen stereophotogrammetry utilizing tantalum indicators has been in use in Lund, Sweden, and by 1977 more than 3000 stereo roentgenograms have been evaluated with that system. Patient investigations started March 1973, and by March 1978, 200 patients have been marked with tantalum indicators and investigated for longer or shorter periods with the method. The reasons for the investigations have been very variable but growth disturbances, craniofacial anomalies, disorders of the spine and visceral tumours have dominated. The aim has been to evaluate the effects of medical and surgical treatments. The high technical reliability and versatility of the method has led to accurate and easy to interpret results with a distinct clinical significance.

A problem in evaluating bi-plane radiographs is the determination of corresponding points. If the angle of convergence is small, this can be done by stereoscopic vision. In the case, when this angle is large and no stereoscopic vision is possible, epipolar rays may be used for supporting the image evaluation. The proposition of epipolar ray correlation is: corresponding image points lie on con-jugate epipolar rays emerging from centers in each of the two image planes. The conjugate rays intersect in the intersection line of the two image planes. For the construction of the epipolar rays different methods can be used: 1) - Mechanically, the epipolar rays can be represented by two threads, one for each radiograph. The correspondence of the threads, also when moved over the images, is established by a simple gear system. An apparatus based on this concept is shown. With this apparatus evaluations of radiographs can be done point by point. 2) - Electronically, the epipolar lines can be plotted by a computer on transparent paper, so that by contacting the radiographs with the drawn lines, again a point by point evaluation is possible. 3) - The epipolar rays can be generated and superimposed to the radiographs by means of the display-stereocomparator described in the paper of E. Hierholzer. As a stereoscopic vision is not possible in most cases of bi-plane radiographs, this device has to be used monocularly.

To objectively and quantitatively demonstrate regional differences in brain endocast morphology, traditional anthropometric caliper measurements must be replaced by a system providing not only localness, but homology and reasonable freedom from allometric distortion. Stereoplotting the radial distances from endocast surface (the closest point to the once underlying brain cortex) to a homologous center every ten degrees provides some 300+ data points for each dorsal endocast surface, thus giving the requisite localness. These measurements provide a large matrix of data suitable for a number of multivariate statistical techniques, and the translation of such data and analyses to readily visualized maps, which can then be compared in relation to both taxonomic and functional knowledge about the cerebral surface. This paper descri-bes some preliminary results from using such methods on a sample of 64 undistorted endocasts composed of both pongids and fossil hominids. While sample sizes within taxonomic groups need to be augmented, the preliminary and tentative pilot studies conducted so far suggest that the method has excellent potential, and that two major areas of the brain endocast surface show the greatest shape changes : 1) the posterior association areas (inferior parietal lobule); 2) the anterior prefrontal areas.

In this survey a review of several investigators' work on computational graphics of neuronal shape and form is presented, and is followed by a proposal that the functional inferences of stereomorphology lead to a three-dimensional integration of brain structure and function. The concepts are, of course, also applicable to the spinal cord, and in some ways to peripheral nerve and muscle. Following a description of the parameters and techniques used to describe neuronal body, dendrite, and branching geometry, a brief account is given of the importance of biostereo-metrical techniques to further understanding of the asymmetry of the brain as a whole. Brief statements of a descriptive nature are provided to indicate studies on stereomorphology of the brain vasculature, and a new technique called cinemorphology which, combined with computed tomo-graphy, allows three-dimensional unfolding views of the whole brain or any structure within it. The latter views can be rotated for additional insight into the structures total relations within the brain.

A primary aim in the physical rehabilitation of individuals with severe head or spinal injuries resulting in hemiparesis, tetraparesis or paraparesis, is the restoration of the functional motor abilities controlling involved muscle groups. The regaining of trunk postural stability provides a valuable antecedent in the recovery of use of the arms and legs. Accurate objective infor-mation must be provided to the therapist for assessment of treatment regimes. At present, few objective practical methods are available to furnish this evaluative information. Therefore the purpose of this study was to investigate the use of a biostereometric range of motion sensor for recording and quantifying trunk static equilibrium in individuals ungergoing therapy for head trauma. The sensor located the relative position of the C-7 vertebrae of the patient in space using continual monitoring of spherical coordinates. Results of the test protocol included : plots of the movement of the trunk excursions, determination of the maximum area of excursion and a trunk sway index (the relationship of the maximum area, the total excursion distance and a time factor). Further results demonstrated that the biostereometric sensor yielded quantitative documentation of improvement in a patient undergoing therapy.

I very much appreciated the invitation to contribute a paper to this Symposium on Applications of Human Biostereometrics, as it provides a valuable opportunity for me to take a fresh look at a problemâ€""the cerebral localisation of psychological function"â€"in which I have been interested for many years. This interest grew out of considerations of the clinically important problem of how we should go about the task of relating the form of the changes in human behavior consequent upon damage to the human brain following, say, head injury, to the form of the changes in brain morphology which constitute that damage, and related issues.

Stereological method on cells and network analysis of dendritic arborization have been used, allowing us to obtain quantitative informations on cytoarchitecture of adult cat inferior colliculus. Moreover these methods allowed a plain change to be objectively established in the cortical afferent region with the degree of functional maturation.

Digital photogrammetry provides a convenient and inexpensive means for determining the threedimensional geometry of the optic nervehead from stereophotographs of the ocular fundus. SR l's photogrammetry algorithm, STEREO, measures the relative height of each point of the ocular fundus by correlating the right and left images of the stereopair. The correlated regions are expanded and contracted in a dynamic fashion to increase the statistical confidence in the height estimate and avoid edge blurring, respectively. The resulting height map is then smoothed with a shrinking operator to remove obviously anomalous structure. Next, colorimetric and densitometric procedures extract the disc boundry in a repeatable way. Finally, the depth and volume of the optic cup are measured with a planimetry algorithm.

There is a common opinion among eye specialists and opticians that children's glasses often are not shaped for optimal fitting. A fundamental reason for this is the lack of data for the shaping of the bows, with the result that most children's glasses are reduced copies of adult's glasses. This report describes a photogrammetric method for collection of primary data for manufac-turing bows for children. An ordinary amateur camera was equipped with a stereo-adapter. With a few arrangements, such as projecting a pattern on the face and keeping the hair away from the ears, 600 children were photographed. A calibration photograph was exposed at the beginning and end of each film or when the equipment had been transported or otherwise disturbed. The photographs were measured in a stereocomparator and the coordinates analytically corrected for distortion. After determination of model coordinates the requested geometric information, such as pupillar distance, eye-ear distance, location of the bridge of the nose etc, was calculated. The shapes of average noses were presented as profile plots.

Laser contour angiography is a new technique which was developed by the author for topographically mapping the retina and optic nerve head of the eye. A striped pattern of laser light is projected onto the ocular fundus through one side of the patient's dilated pupil. Photographs, taken through the opposite side of the pupil, reveal the topography as parallactic displacements of the stripes. A topographic map is constructed from tracings of the stripes. The method was shown to produce highly reproducible maps. In order to further improve the reproducibility of the topographic measurements, the effects of the man-machine interaction during the stripe tracing process on measurement error were investigated. The tracing strategy of the operator is to estimate the position along the edge of a stripe at which the perceived light intensity is halfway between the local maximum and minimum. The relationship of errors in the estimates of position to those in perceived intensity was established using a theoretical model. The predictive ability of the model was tested using repeated measurements on the laser contour angiogram of a human eye. The model predicted a position error (standard deviation) of 4.3 micrometers. This agreed remarkably well with the experimental result of 4.5 micrometers. The results indicated that film grain noise and defocussing of the projected pattern are the primary contributors to measurement error. We now use a finer grain film, and have further reduced the position error to 3.4 micrometers.

Glaucoma is a disease characterized by increased ocular pressure with optic nerve atrophy and loss of visual field or side vision which eventually leads to blindness. The surface of the portion of the optic nerve within the eye, the optic disc, becomes depressed and its contours are altered. The optic disc becomes more cupped with increasing progress of the disease. For measurement of the cup, retinal cameras can provide simultaneous stereo photographs of the optic disc taken through the pupil. However, the photographs have to be obtained at a narrow angle due to the limit of the size of the dilated pupil. Our studies of the geometry of the stereo photographs show a greater error in depth compared to horizontal dimensions because of this biological limit. Similarly, data for the optimal conditions for photography of the optic disc, including aperture size and magnification have been obtained. Since the changes in the optic disc can progress with increase of the desease, photogrammetric techniques for measuring relative changes in the dimensions of the cup have been developed. Both subsequent and initial photographs are measured after being similarly oriented in space. The photogrammetric measurement of changes in the optic disc cup in glaucoma has potential as a useful technique which can have wide clinical application.

A digital photogrammetric method capable of providing quantitative measurements from stereophotographs of the optic nerve head has been utilized in our laboratory to study the optic cup in patients with ocular hypertension and chronic glaucoma. Studies on the reproducibi-lity of the digital photogrammetric technique of optic nerve head photographs taken with various cameras has revealed that (1) reproducibility was significantly better when simultaneous stereo-photographs were analyzed and (2) the Donaldson stereofundus camera provided the best reproducibility of the majority of geometric measurements in optic cups of various sizes. In addition, it has been determined that the shallower cups encountered larger percentage errors. Improvements in parameter reproducibility have been obtained when finer sampling intervals have been used during film digitization. An additional study comparing reproducibility of cup parameters as determined by the digital photogrammetric procedure with those values obtained from analog photogrammetry and manual analysis revealed generally superior reproducibility of the digital photogrammetric analyses in the eyes studies. Preliminary observations utilizing the digital photogrammetric procedure in a modified Zeiss model eye indicated that minor shifts in camera posi-tion did not alter the reproducibility of the technique.

A method is presented for quantitatively expressing the topography of the human optic disc, applicable in a clinical setting to the diagnosis and management of glaucoma. Pho-tographs of the disc illuminated by a pattern of fine, high contrast parallel lines are digitized. From the measured deviation of the lines as they traverse the disc surface, disc topography is calculated, using the principles of optical sectioning. The quantitators applied to express this topography have the the following advantages : sensitivity to disc shape; objectivity; going beyond the limits of cup-disc ratio estimates and volume calculations; perfect generality in a mathematical sense; an inherent scheme for determining a non-subjective reference frame to compare different discs or the same disc over time.

Studies on the somatic growth of young children have traditionally been made using conventional anthropometry techniques. As a result, while the conditions of growth of morphological variables such as weight or segmental dimensions are well known, the same cannot be said of the more global aspect of the development of the body in a three-dimensional reference space. Yet body volumes and surfaces represent morphological characteristics which are just as necessary for a good understanding of physiological phenomena (thermoregulation, energy balance, etc.) as the conventional linear data. In the volume of their research on children's growth in recent years, the authors have found that in none of the studies mentioned in the literature was consideration given to the dynamic aspect of the child's somatic development in a three-dimensional space. A primary reason for such omission is largely to be found in the technical difficulties encountered in the measure-ment of somatic characteristics such as body volume and surface. Yet, among the several possible methods of study, biostereometry and particularly the photogrammetric tool, is certainly one of the most rewarding. This being so, the authors propose to use the photogrammetric technique to undertake, in a first stage, a methodological study that will draw up, on a limited sample of infants and young children, the development chart, over a period of time, of the surfaces and volumes of segmental elements. Thus will be checked the relationships between the growth rates of different characteristics : surfaces, volumes, weight, linear dimensions. Quite apart from the intrinsic value of such studies, the data thus collected will eventually provide practitioners, pediatricians and physiologists with the reference records that have so far been lacking.

A system is described for the recording and measurement of human body shape by a series of circular ultrasound scans. Computer manipulation of the echo data provides a graphic display of body contour, and a measurement of total body surface area and volume. The theoretical resolution for distance measurements using this device is 2.5 mm, a figure achieved in practical calibration experiments using a metal test object. Measurements from the body surface, although less precise, are sufficiently accurate and reproducible to enable useful clinical information to be obtained, particularly in recording the morphological changes associated with obesity and malnutrition.

We are taking it as a starting point, that a living body could never occupy the same space determined by three axis. As consequence biostereometrics has to operate with all-round informations. The non contact, all-round biometry of a freely modving, living body is realized by different researchers. It is actual to determinate the proceedings with spatiotemporal informations. Studying the breathing movements of the thorax and abdomen, the best method is to compose the actual Volume Distribution Profile, VDP, of the trunk. Produced in a regular frequency VDP-s can be applicated on the time axis and so we compose the Integrated Surface of the Breathing Movements, ISBM, which is characteristical for age, sex and type. With a computed system it is easy to visualize on a display the form-changing of VDP-s during breathing movements. We hope, that the time isn't far way when we can use this parameter as the ECG. This is a global information. Localized informations can be produced also by means of horizontal sections. The form-changing surface of identical sections during the breathing gives localized data. One of the next steps must be a data reduction, to obtain minimum number of relevant points, the help of which patients can be grouped.

Biostereometric analysis of body form was performed preflight and postflight on the 9 Skylab astronauts, using 4-camera stereophotogrammetry. The estimated standard deviation of the measurement of total body volume was 2-3%, but it should be possible to improve this to 1-2%, given some comparative studies with other methods, and some improvements in technique. The accuracy would then be comparable with underwater weighing, which is acknowledged to be the best currently available method for determining body density, and hence calculating body fat. Improvements beyond this accuracy are unlikely to lead to an enhancement in the accuracy of estimating body fat, because of uncertainties in the relationship between body fat and body density. Biostereometric analysis possesses one important advantage over underwater weighing, in that it is able to measure the volume of any chosen region of the body. This ability was used on Skylab to study the changes in leg muscle over the course of the flight, and also to measure the changes in body fat, which were particularly evident in the volume of the abdomen and buttocks. This approach would be greatly enhanced, however, by studies on the effects of diet and exercice on the volume of different body regions.

A method is presented for easy spatial reconstruction of the complete human thorax. Eight cameras are used to photograph the subject simultaneously. Position and orientation of the cameras is not critical and no alignment tool is needed as the use of a reference frame allows easy computerized determination of these parameters a posteriori. The eight photographs are digitized. Processing of the digitized data by computer results in the determination of the spatial coordinates of any arbitrary point on the thorax (such as characteristic bodymarks, electrodes, etc...). After axistransformation and representation in cylindrical coordinates, smoothing of these coordinates is realized with a special computer program using bicubic B-spline functions. Mathematical constraints for this program are provided by a set of anthropometrical measurements. In this way the complete thorax is reconstructed and an analytical function of the thorax surface is obtained. The model of the thorax, obtained with this cinematographical technique is validated by the results of axial tomography.

A facial contouring technique, using light sectioning by Coob was modified by Ainsworth and Joseph and used in a numerical study of children with isolated pulmonary stenosis (PS) to test the hypothesis that the facial pattern in this condition differs from the normal. Measurements were compared between a group of 20 normal children, and a group of 20 children with PS between the ages of 6 and 10.5 years. A distinctive facial pattern has emerged. Many anteroposterior measurements were significantly greater in the PS group, indicating that the tissues are more prominent in the maxillary region. Twenty-nine of the measurements showed significant differences between the two groups (P <.05). Discriminant analyses were carried out to discover which, if any, might be used to predict the group to which an individual should belong. Depending on the variables chosen, between 34 and 37 individuals from the total of 40 were assigned to their correct group, PS or control.

In order to determine the part played in facial dissymmetry observed on a living person by the various constitutive elements of the cephalic tip (the soft parts - skin, muscles and the underlying bone structure) we undertook, using a biostereometric method, to evaluate asymmetries between homologous right and left dimensions on a living person's face and on a skeleton. While in an individual, a marked degree of facial dissymmetry can sometimes be observed; average differences between the right and left sides of the face may nethertheless balance out, and remain slight. Conventional anthropometrics techniques do not show up such slight values. With a view to securing a higher degree of accuracy, study of the stereometric technique of measurements. Using this technique, quasi imperceptible differences between the right and the left sides of the face on a living person as well as on a skeleton, together with variations in the orientation or angulation of anatomical segments in a three-dimensional space can be measured. We were thus able to detect, in a number of dry skulls, average differences of approxi-mately a millimetre between the two sides of the face which cannot be attributed to back of accuracy in measurements. Although statistically the difference are not always significant, the para-metric values of facial dimensions are invariably greater for the left side. On the other hand, for the sample of living subjects as a whole, the differences between homologous distances are not statistically significant. But it may be that, on a living subject, the experimenter is inclined to take measurements that are susceptible of symmetrization (for instance, the nasion in the median sagittal plane) whereas on a dry skull anatomical reference marks can be determined with the utmost accuracy. It may be inferred from there results that the softer parts tend, as a rule, to correct the dissymmetry of the underlying skeleton.

The aim of this study was to measure serial growth changes of the lips in the human between birth and 19 years of age. The lips were measured three-dimensionally, using short base stereophotogrammetry, in terms of mouth width, mouth height and anterior projection of mid points of upper and lower lips relative to a plane joining the angles of the mouth. Lip shape was measured in angular terms related to these parameters. The bulk of the findings was derived from 112 plots from series for 6 pairs of like-sexed twins between the ages of 9 and 19 years, but the results for one child were included to show how the rapid contrasting growth changes of the baby and infant merge into the pattern for older children.

The initiation of a research project involving stereophotogrammetry for the investigation of changes in facial morphology in children over a period of ten years (Burke P.H.(1972). The accuracy and range of certain stereophotogrammetric measurements of facial morphology. Transactions of the European Orthodontic Society) led to the establishment of Biostereometrics in Addenbrooke's Hospital. This is now used for recording and studying structural changes due to growth, disease or trauma as an aid to diagnosis and treatment of patients. The acceptance of this technique has been largely due to the adherence to pure photogrammetric principles which are easily understood and to good photography leading to simple and rapid analysis of the stereopairs. The equipment and methods used together with examples of ongoing projects will be described.

Man's oral-facial structures are vital for the functions of breathing, mastication, swallowing, vision, and communication. When defective development of these tissues occurs, function becomes impaired and the anatomic features of the afflicted individual will frequently deviate from the norm. This error of form and function will classify the individual as being physically and psychosocially handicapped. The successful habilitation regimen of the handicapped person depends on the accurate analysis of both craniofacial anatomy and physiology of these individuals, as well as psychological implications of the disfigurement. Biostereometrics can contribute to the establishment of operationally valid measures for assessing the severity of the handicapping conditions. The heterogeneous nature of diverse disfigurement suggest that an improved classification of malformations would be beneficial. Three-dimensional analysis may also have significant influence on the accuracy of the diagnosis, and the establishment of a biologically sound treatment plan. Biostereometrics will contribute more fully if the three-dimensional surface analysis can be coordinated with a study of 1) the underlying skeletal structures, and 2) the operational musculature. Increased communication between the stereometric experts and the biological scientists should accelerate the application of this technique to the health problem.

Studies of the systematic relationships within the primate taxa Homo, Hylobatidae (lesser apes), Hominoidea (apes and man) and Colobinae (Asian and African leaf monkeys) are described. All are based in large part on multivariate statistical analyses of cranial morphology. Adequate quantification of the frequently complex and subtle differences in the morphology of the animals being compared, as well as the inclusion of statistically adequate samples of as many presumed species or other groups of interest as possible, are essential to the success of such analyses. Two methods of stereometric measurement have been developed to make this possible. In initial studies of the Hylobatidae and the Hominoidea, a simple mechanical device was designed which determines the tri-dimensional coordinates of an anatomical point by measuring an angle and two distances. An improved version was used in an investigation of Subsaharan human crania. In a taxonomic revision of the Colobinae now in progress, crania are photographed in several views with a pair of metric cameras; point coordinates are then measured in a modified stereoplotter and the views rotated mathematically into a single coordinate system. Although stereometrics is only one component in a complex system of analysis, it is an extremely important one. Taxonomic revisions of the described scope and depth could not be carried out with conventional methods of measurement without a much greater commitment of resources, if at all.

The scanning electron microscope (SEM), combining high resolution and large depth of focus, affords detailed observation of surface microstructure in a three-dimensional perspective. It also allows large specimen dimensions and avoids the processing and sectioning limitations of light and transmission electron microscopic procedures. For these reasons the SEM is ideally suited for analyses of bone, a rigid tissue whose surface topography and internal architecture accurately reflect the developmental, metabolic and mechanical influences exerted upon it. Furthermore, SEM photomicrographs are compatible with devices for quantification, mathematical manipulation and graphic reconstruction of the image. Features of a photo may be traced with a stylus on the electromagnetically activated surface of a data digitizer, which converts the outlined path to x and y axis coordinates. Interfaced with a programmed calculator these data undergo algebraic and geometrical computation and may be stored for statistical analyses. Alternatively, stereopairs of micrograph transparencies may be utilized in micro-stereophotogrammetric procedures in which x, y and z axis coordinates are generated for selected morphologic points. Our research concerns spatiotemporal interrelationships of primate craniofacial growth as evidenced by changes in the skeletal gross morphology and microanatomy of the orbital region, jaws and teeth during their growth and development. Applications of SEM and digitization techniques to these studies and an evaluation of the derived data will be presented.

In spite of a long and rich history of investigation, numerous questions remain to be resolved about human form and form change. Techniques from analytical geometry, anthropometry, biostatistics, comparative, developmental and functional anatomy, osteology, paleontology and scanning electron microscopy are but a few of the methods which have been used to advance our understanding of human anatomy. Biostereometrics, however, with its capacity to generate mathematically precise three-dimensional descriptions of body structures has not been broadly applied in comparative studies of human form and morphogenesis. This paper briefly identifies and reviews some of the more persistent problems associated with the description, quantification, and analysis of biological form; it offers suggestions how biostereometric procedures may be combined with other methods to generate more complete, mathematically accurate and biologically sound characterizations of human form, growth and development.

We have recently entered a period of heightened awareness about the significance of human body geometry at all levels of organization from the subcellular to the organismic. In particular, the study of subtle variations in the irregular three dimensional form of the heart, the brain, and the spine has attracted substantial research interest. This trend has been amply demonstrated in the papers presented over the last three days. Several of the reports provide insights which point the way for many years of further research. The wide ranging scope of the symposium makes it difficult to summarize, but perhaps an attempt should be made to document some of the more important developments leading up to this meeting and, also, see if we can predict where biostereometrics is going.