Grady Early Forensic Anthropology – Researching Methods for 3D Scanning Femora for Use in Stature Estimation
Introduction
My internship was with the Forensic Anthropology Center at Texas State (FACTS) under the supervision of Dr. Deborah Cunningham. My primary role was to conduct research on the 3D scanning of femora to gather pilot data to support research on gathering new stature estimates from Homo erectus, one of our early ancestors. Our research was conducted in the Grady Early Forensic Anthropology Research Lab (GEFARL) and aimed to test the effectiveness of stature estimation through 3D scanning. Our research project has three objectives: to provide applicable pilot data to support research on Homo erectus stature, to offer procedures to efficiently scan femoral data, and to create reliable 3D models to support fellow researchers.
Homo erectus is named for its ability to stand up and walk. This species is believed to be the oldest early human with a human-like body. Homo erectus is considered one of our direct ancestors and by further understanding this species we can further understand the past, present, and future of human evolution. New insight on how the body changed during human evolution has created more questions for the true stature of Homo erectus (Graves et al., 2010). Stature estimation aids in the understanding of basal metabolic rate, energy expenditure, diet, locomotion, and provides for implications for the evolutionary trajectory of the genus Homo. Current issues, however, complicate body size estimates for early hominins. These complications include small sample sizes and inaccurate femoral length estimations caused by fragmentary femora of hominin specimens (Anton, 2007). This provides cause for problematic interpretations of body size and its relation to Homo evolution.
Stature is an individual’s natural height and is traditionally estimated through measurements of the whole femur, the thigh bone. Because of this, stature data on individuals with broken femurs is not typically be included in samples. This project is a two-part study. The first is to develop a method to efficiently collect a 3D sample of the Texas State Donated Skeletal Collection through laser scanning. The second is to compare measurements of the 3D sample to the physical measurements of the femora that were scanned in order to test the accuracy of the 3D models. My research thus far has focused on the first part. A large part of my internship was to research information on how to use the Next Engine Scanner and how to incorporate it into my study.
Incorporating methods to create 3D models of the femur was difficult at first. This was because researchers that used 3D femoral data only vaguely described their procedures in gathering a 3D sample. As a result, we decided that a goal of this project would be to make it easy to replicate.
We are using this method because 3D data is efficient and more cost effective than traditional methods of gathering osteological data. 3D data will not replace the information that comes from a physical bone sample however its relevance comes from how efficient it is to distribute and access. The limitations from a physical sample of bone come from samples being located in a central location. 3D data alleviates the expenses for traveling to a sample or the risk of shipping a fragile sample to a different research facility This research attempts to provide evidence for the accuracy of 3D data for femoral length estimation.
Materials and Methods
Using the Texas State Donated Skeletal Collection, the Next Engine Scanner was incorporated to collect 3D images. This was in order to test the accuracy of segment-based stature estimation through 3D models.
Fellmann’s Method (2006)
Fellmann’s method is a technique to measure the stature of Homo erectus through broken femora. We used this system to provide context while creating 3D models by paying close attention to the landmark data that would be needed for the future comparison of data. This method of estimating stature through fragmentary femora incorporates a segmentation-based regression system to interpret the femoral length.
Femoral length is estimated using a correlation between the lengths of 4 sections throughout the femur. The first phase begins at the most proximal point of the femur on the most superior aspect of the femoral head and ends at the midpoint of the lesser trochanter. The second phase begins at this midpoint and ends at the point where the supracondylar lines of the linea aspera are no longer parallel on the shaft the femur. The third phase begins at this point and ends at the most superior point of the intercondylar fossa between the distal articular facets. The fourth phase starts at this point and ends at the most inferior point on the medial condyle.
The method finds patterns between the length of each phase and uses those to estimate the length of the whole femur. The study includes fossils from different geographical locations and provides different regression formulas to estimate femoral length for each group. When femoral length is estimated it can be used to estimate stature using traditional methods. Essentially, the length of each phase can estimate the total length of the femur. This study reveals the relationship each phase has with each other, in regard to total femur length. The estimated femoral length is then used to estimate stature using standard methods for estimation stature.
The Next Engine Scanner
We used The Next Engine Scanner because of its .005 inch accuracy and its relative simplicity when compared to other scanners. The Next Engine Scanner is a desktop 3D surface scanner that we are using to scan the femora of the willed body collection. The scanner works by creating 3D models using data collected through lasers as input. The item to be scanned is placed on a device called an auto drive. The auto drive is the tool that rotates the item 360 degrees. The basic settings per scan vary depending on the part of the bone we are measuring. These settings include the type of scan, the divisions, and the focus of the scan (Sqyures, 2016).
The program allows for three different types of scans: the 360 scan, the bracket scan, and the single-shot scan. The 360 scan allows the item to be scanned on a complete 360-degree rotation; the bracket scan lets an individual scan an item on a partial rotation that can be adjusted; lastly, the single-shot scan creates a 3D model of a single division.
Divisions are jargon for how many individual scans are wanted to create the 3D scan. For example, a 360 scan with 16 divisions will take 16 scans around 360 degrees of the item to create the 3D image; a bracket scan with 3 divisions, however, will only take 3 scans surrounding only a partial rotation of the item.
The focus of the scans was macro, wide, and extended. This refers to the size of the item. Smaller items that would require a higher focus would use the macro setting, while items with a larger surface area would use the wide setting.
Texas State Donated Skeletal Collection
The sample that I scanned belonged to the Texas State Donated Skeletal Collection at the Forensic Anthropology Center at Texas State. When individuals are donated to FACTS they pass through several research labs before their curation. First, they are brought to the Forensic Anthropology Research Facility (FARF) where decomposition and search and recovery of human remains are studied. Then they are processed at the Osteological Research and Processing Lab (ORPL). This lab is used for processing remains for curation into the Texas State Donated Skeletal Collection, located in GEFARL. GEFARL also has equipment for 3D modeling and histology. More information on the labs of the FACTS can be found here.
Methods for Scanning
This part of the project involved a great deal of trial and error. Originally, we wanted to incorporate a device called the multi drive. The multi drive is a tool, similar to the auto drive, that tilts the object to allow for more accurate data. The multi drive is only compatible with the macro scan which created complications on the meshing of the 3D data. Initially, the objective was to incorporate 5 scans, however scans of the shaft of the femur would not align. When I attempted to scan the midshaft of a femur, the scans would auto-align to create an inverted image of the femur.
Instead, I took four separate scans of the femur using the auto drive. These scans were taken from the distal half, the distal articular surface, the proximal half, and the proximal articular surface. The scans of the distal half and the proximal half consisted of 360 scans using 16 divisions. The distal articular surface and the proximal articular surface used a single-shot scan. I placed the femur on the auto drive and anchored the bone with Velcro straps, so it would not fall during its rotation.
Before I took the scans, I placed reconstruction markers on the femurs to more efficiently align the four scans together. A total of nine reconstruction markers were placed on the femur in three different areas. I created these markers by hot-gluing airsoft pellets to areas of the bone that would help with the virtual alignment of the bone without affecting features of the femur. To connect the distal articular surface scan with the distal half scan, I placed three reconstruction markers on the distal articular surface. These were put on the lateral condyle, the medial condyle, and the patellar surface. To connect the scan of the distal half with the scan of the proximal half, I then placed three reconstruction markers on the midshaft of the femur. I placed these markers on both sides of the linea aspera and the anterior aspect. In order to find the location for the midshaft reconstruction markers, I measured the maximum femoral length of each individual I scanned and divided the measurement in half to locate a midpoint. To connect the scan of the proximal half with the scan of the proximal articular surface, I placed three reconstruction markers on the proximal end of the femur. I then placed two markers on the anterior and posterior aspect of the proximal head and the third was placed on the superior aspect of the greater trochanter.
Results
I collected 3D data from the femora of individual and then aligned the scanned data to create 3D models. Here are the results of scanned data.
Conclusion
The aim of this research is to provide a systematic procedure for additional scans to be taken in an efficient way that is easily recreated. In future study, scans of up to 40 individuals would be taken from the Texas State Donated Collection and be compared to physical data of the same sample. The femora of ten individuals from each of the groups of White Males, White Females, Black Males, and Black Females will be scanned in order to test for geographical variability. The comparison will be of measurements taken from the four segments of Fellman’s method (2006). We are comparing these measurements to test if segmentation regression on a 3D model is a viable option for stature estimation.
The data that is collected would then be used to support future studies with stature in Homo erectus. Dr. McCarthy proposes to travel to museums to collect 3D models of Homo erectus samples. He and his colleagues then plan to create femoral length estimates from the fractures remains using Fellmann’s method. With the estimated femoral length, they then plan to create more accurate stature estimates for Homo erectus. These stature estimates could then be used to redefine our understanding of human evolution.
This internship has been an amazing opportunity to give me a better understanding into my future in anthropology. Dr. Cunningham has been a great mentor in pushing me to consider graduate schools and to offer insight to researching a graduate program that would best suit my interests. This internship has also offered me a great number of opportunities. I was able to volunteer for the American Association for Physical Anthropologists (AAPA) 2018 annual meeting where I got to network with researchers relevant to my interests. It was in Austin and I got to be a photographer and make a lot of people uncomfortable while asking them if I could take their picture. I was also able to present my research at the Undergraduate Research Conference at Texas State University. A goal of mine with this internship is to participate at next year’s AAPA meeting in Cleveland, Ohio by participating in the Undergraduate Research Symposium. Being able to present part of my research helped me gain some experience in presenting a poster and clearly explaining my research. Most importantly, I was able to work near the professors studying Biological Anthropology. This enabled me to ask them for reference letters when applying for opportunities for this summer. Because of this internship, I will be interning at the University of North Texas Health Science Center. I am grateful for this internship, even though it was a lot of hard work, it’s helped me become a stronger anthropologist.
References
Antón S.C., Spoor F., Fellmann C.D., Swisher III C.C. (2007) 11 Defining Homo erectus: Size Considered. In: Handbook of Paleoanthropology. Springer, Berlin, Heidelberg
Fellmann C. D. (2006) Estimation of femoral Length and Stature in Homo erectus from Fragmentary Remains. Department of Anthropology, New York University, New York, New York.
Graves R. R., Lupo A. C., Mccarthy R. C., Wescott D. J., Cunningham D. L.. 2010. Just how strapping was KNM-WT 15000? Journal of Human Evolution 59:542–554.
Squyres N. S. (2016) Shape Variation in the Distal Femur of Modern Humans and Fossil Hominins. Johns Hopkins University, Baltimore, Maryland.