Photogrammetry for 3D digitizing bones of mounted skeletons: Potential and limits

Since the late 20th century, new technologies have provided powerful ways to digitize biological structures in three dimensions (3D). Among those, photogrammetry is a low cost and non-destructive method, which has become increasingly used since the development of the digital camera. Recent studies have demonstrated that reconstructions of isolated elements can be of as high quality as those obtained with laser scanners. Here, we wanted to test the performance of photogrammetry for the quantitative analysis of mounted specimens in museum exhibitions. Indeed, access to material can be an issue in comparative anatomy and, especially, in paleontology. This is notably the case for large, impressive specimens. We performed reconstructions based on acquisitions done under various conditions and also tested the reconstruction performance of two software programs. The resulting 3D models were then compared to a reference object corresponding to the bone of interest digitized with a cutting-edge surface scanner. Our results show that photogrammetry enables quality reconstruction of the almost entire surface of the mounted bone of interest. Photogrammetry thus appears a reliable method perfectly suited to study large specimens exposed in museum gallery.     

A new stereophotographic technique for analyzing the three-dimensional structure of fish schools

Synopsis A technique using two downward-directed 35 mm cameras has been modified to measure the three-dimensional structure of fish schools. The resulting stereo pairs of photographs are analyzed, producing the 3-coordinate location of each fish's nose, after correction for lens distortion and refraction. Separation angles (bearing and elevation) and distance can then be determined for any pair of fish in the school. The technique's high level of accuracy is demonstrated for an underwater calibration field. It is then applied to the measurement of the 3-D structure of schools of coho salmon (Oncorhymchus kisutch) swimming in a hatchery trough. Although the fish were not organized in a rigid crystal lattice, the analysis provided some evidence of structure.     

Computer tools for the analysis of schooling

Synopsis Quantification of fish school structure is difficult because of: (1) the large amount of positional data that must be recorded, (2) the fact that schools are moving, and (3) the fact that schools are threedimensional. Computer tools for addressing these problems include x,y digitizing devices interfaced to computers and interactive computer graphics systems. General computer algorithms now exist for 3-D reconstruction of schools given any two 2-D views such as photographs.Our computerized film analyzer, the Galatea system, is described in detail. This system allows for rapid, accurate input of multiple points which can be followed through each frame in a film sequence. Its software packages perform 3-D reconstructions, analyze undulatory motion, and do extensive statistical tests.     

A new geometric-based model to accurately estimate arm and leg inertial estimates

Segment estimates of mass, center of mass and moment of inertia are required input parameters to analyze the forces and moments acting across the joints. The objectives of this study were to propose a new geometric model for limb segments, to evaluate it against criterion values obtained from DXA, and to compare its performance to five other popular models. Twenty five female and 24 male college students participated in the study. For the criterion measures, the participants underwent a whole body DXA scan, and estimates for segment mass, center of mass location, and moment of inertia (frontal plane) were directly computed from the DXA mass units. For the new model, the volume was determined from two standing frontal and sagittal photographs. Each segment was modeled as a stack of slices, the sections of which were ellipses if they are not adjoining another segment and sectioned ellipses if they were adjoining another segment (e.g. upper arm and trunk). Length of axes of the ellipses was obtained from the photographs. In addition, a sex-specific, non-uniform density function was developed for each segment. A series of anthropometric measurements were also taken by directly following the definitions provided of the different body segment models tested, and the same parameters determined for each model. Comparison of models showed that estimates from the new model were consistently closer to the DXA criterion than those from the other models, with an error of less than 5% for mass and moment of inertia and less than about 6% for center of mass location.     

Body segment inertial parameters and low back load in individuals with central adiposity

There is a paucity of information regarding the impact of central adiposity on the inertial characteristics of body segments. Deriving low back loads during lifting requires accurate estimate of inertial parameters. The purpose was to determine the body segment inertial parameters of people with central adiposity using a photogrammetric technique, and then to evaluate the impact on lumbar spine loading. Five participants with central adiposity (waist:hip ratio>0.9, waist circumference>102 cm) were compared to a normal BMI group. A 3D wireframe model of the surface topography was constructed, partitioned into 8 body segments and then body segment inertial parameters were calculated using volumetric integration assuming uniform segment densities for the segments. Central adiposity dependent increases in body segment parameters ranged from 12 to 400%, varying across segments (greatest for trunk) and parameters. The increase in mass distribution to the trunk was accompanied by an anterior and inferior shift of the centre of mass. A proximal shift in centre of mass was detected for the extremities, along with a reduction in mass distribution to the lower extremity. L5/S1 torques (392 vs 263 Nm) and compressive forces (5918 vs 3986 N) were substantially elevated in comparison to the normal BMI group, as well as in comparison to torques and forces predicted using published BSIP equations. Central adiposity resulted in substantial but non-uniform increases in inertial parameters resulting in task specific increases in torque and compressive loads arising from different inertial and physical components.     

New anthropological data from Cussac Cave (Gravettian,Dordogne, France): In situ and virtual analyses of Locus 3

Cussac Cave presents a unique combination of parietal art and several hundred parts of scattered human remains, dated to the Middle Gravettian (29–28,000 cal BP). The cave is protected as a National Heritage site. As a result, only noninvasive bioanthropological analyses are allowed, consisting of in situ observations and the study of 3D models obtained by photogrammetry. Here we present the first results of these analyses of the human remains from Locus 3. Only 65 of the 106 human skeletal fragments and bones could be firmly identified. Virtual analyses were carried out on 3D models of 16 skeletal elements so that osteometric data could be provided. Despite the limitations inherent in studying commingled remains and those specific to Cussac Cave, the search for virtual pair-matching, articular congruence, and osteometric sorting allowed the allocation of twelve bones to three individuals, one late adolescent and two adults.     

Morphometric identification of individuals when there are more shape variables than reference specimens: a case study in Galápagos tortoises

Molecular biology techniques are useful for taxonomic assignment, but they are not always accessible and can be expensive and time consuming to perform. Morphological methods to identify the origin of individuals could be valuable if they can be performed rapidly, accurately, and with minimal resources. In order to correctly assign the origin of individuals from two distinct tortoise lineages, we studied here the accuracy of shape statistics depending on the inclusion of different numbers of shape components. Misleading assignment may occur if an optimal balance between the number of shape variables and the number of sampled individuals is not respected, especially when more variables than specimens are available. Assignment of museum samples of unknown origin suggests that they mostly belong to only one of the two lineages.     

Seasonal Growth Rate of the Sponge Haliclona oculata (Demospongiae: Haplosclerida)

The interest in sponges has increased rapidly since the discovery of potential new pharmaceutical compounds produced by many sponges. A good method to produce these compounds by using aquaculture of sponges is not yet available, because there is insufficient knowledge about the nutritional needs of sponges. To gain more insight in the nutritional needs for growth, we studied the growth rate of Haliclona oculata in its natural environment and monitored environmental parameters in parallel. A stereo photogrammetry approach was used for measuring growth rates. Stereo pictures were taken and used to measure volumetric changes monthly during 1 year. Volumetric growth rate of Haliclona oculata showed a seasonal trend with the highest average specific growth rate measured in May: 0.012 +/- 0.004 day(-1). In our study a strong positive correlation (p < 0.01) was found for growth rate with temperature, algal biomass (measured as chlorophyll a), and carbon and nitrogen content in suspended particulate matter. A negative correlation (p < 0.05) was found for growth rate with salinity, ammonium, nitrate, nitrite, and phosphate. No correlation was found with dissolved organic carbon, suggesting that Haliclona oculata is more dependent on particulate organic carbon.