| Abstract|| |
Aim: The present study was conducted to determine the most accurate bite mark overlay fabrication technique by studying two physical characteristics, i.e., area and rotation of biting edges of anterior teeth of thirty volunteers. The objective of the study was to evaluate the reliability and efficacy of five commonly used methods of human bite mark overlays using two dimensional (2D) digital images of dental study casts as a gold standard, to rank different methods according to statistically based determination of relative accuracy of each method and to determine its feasibility in Forensic science.
Materials and Methods: Overlays were produced from the biting surfaces of six upper and six lower anterior teeth of 30 volunteers using the following five methods: a) hand tracing from study casts, b) hand tracing from wax impressions, c) xerographic method, d) radiopaque impression method and e) 2D computer-based method. Area of the biting edges of the anterior teeth and relative rotation of each anterior tooth were measured and compared.
Results: The xerographic method was determined to be the more accurate method with respect to tooth area and rotation. Hand tracing methods, from either wax impressions of teeth or directly from study casts, were determined to be inaccurate and subjective.
Conclusions: It is recommended that forensic odontologists discontinue the use of hand tracing overlays in bite mark comparison cases as there is lot of scope for manipulation and observer bias.
Keywords: Bite marks, bite mark techniques, overlays
|How to cite this article:|
Maloth S, Ganapathy K S. Comparison between five commonly used two-dimensional methods of human bite mark overlay production from the dental study casts. Indian J Dent Res 2011;22:493
Bite mark (BM) is considered as a pattern produced by human or animal dentitions and associated structures in any substance capable of being marked by those means.  Bite marks are accepted as being unique to each person even in identical twins since the characteristics of bite mark may be affected by the type, number and peculiarities of the teeth, dynamics of occlusion, muscle function, individual tooth movement and temporomandibular joint (TMJ) dysfunction in the perpetrator. ,,
|How to cite this URL:|
Maloth S, Ganapathy K S. Comparison between five commonly used two-dimensional methods of human bite mark overlay production from the dental study casts. Indian J Dent Res [serial online] 2011 [cited 2019 Sep 19];22:493. Available from: http://www.ijdr.in/text.asp?2011/22/3/493/87079
Bite mark investigation includes the physical comparison of the unknown mark found on skin or objects to known exemplars of the suspect's teeth and metric analysis of suspects teeth. The tooth exemplar, independent of the method used to produce it is called an "overlay" when biting surface data is transferred to clear acetate. This is physically compared to the injury on the skin or a patterned mark. The "hollow volume overlays" records the perimeter of biting surface of each tooth and leaves the inner aspect of the tooth image transparent. There are various methods of overlay production for bite mark analysis. ,,,, The use of digitally produced overlays provides accurate and a more objective method of comparison between the dentition of a suspect and the bite mark. ,,, The computer-based technique is used as a gold standard for bite mark analysis.  In this method, transparent overlays are fabricated from images generated by scanning the suspect's dental casts. When the bite mark image and the overlay are completed, a metric or non-metric analysis can be carried out.
Compound overlay using digital imaging provides information of not only the perimeter outlines of the teeth, but also images of the individual features like, chipping, wear facets, unusual anatomy, and dental erosion within the outlines from the suspects scanned dental casts. Since the bite injury is spatially a three-dimensional (3D) event, it should be ideally analyzed with 3D technology. However, despite the evident need for three dimensional representations of the bite mark, only a few methods have been developed to date. The 2D computer-based technique for analysis of bite mark overlays from the suspect's teeth remains the most common approach as the technology required is more affordable and easily accessible than for the 3D methods.
The present study was conducted to determine the most accurate bite mark overlay fabrication technique by studying two physical characteristics, i.e., area and rotation of biting edges of anterior teeth of thirty volunteers.
The objective of the study was to evaluate the reliability and efficacy of five commonly used methods of human bite mark overlays using 2D digital images of dental study casts as a gold standard, to rank different methods according to statistically based determination of relative accuracy of each method and to determine its feasibility in forensic science.
| Materials and Methods|| |
Thirty subjects with complete set of natural upper and lower anterior teeth were selected for the study. Subjects with orthodontic appliances, intraoral prosthesis, impaired mouth opening, position abnormalities and loss of anterior tooth structure, periodontal and gingival abnormalities, developmental tooth anomalies and severe wasting diseases were excluded from the study.
The upper and lower alginate impressions were taken from each of thirty subjects.  Die stone (Kalrock) models were obtained from each impression and serially numbered. Overlays were produced from the biting surfaces of six upper and six lower anterior teeth of each set according to the technique described by David Sweet et al.  using the following five methods:
- Hand tracing from study casts: A 100-micron thick transparent sheet was positioned over the biting surfaces of each set of upper and lower study casts and the perimeter of the biting surface of the anterior teeth was subjectively recorded by hand tracing using a soft fine tipped felt pen, (Uchida Yoko Co. Ltd., Japan), The orientation mark was added and serially numbered. The overlay was scanned and the image was saved as direct cast tracing overlays [Figure 1].
- Hand tracing from wax impression method: The study casts were pressed into a single wafer of modelling wax sheet (Hindustan modelling wax No.2) to produce a shallow impression of the biting surfaces of the six upper and six lower anterior teeth. A sheet of transparency film was placed over the wax sheet and the perimeter of each of the shallow depressions caused by teeth of interest was hand traced using fine tipped felt pen. Laterality markers were added and numbered. The image on the overlay was assumed to be life-sized because it is generated from a wax impression of the study cast. The overlay was then scanned and saved as wax imprints overlay [Figure 2].
- Radiopaque wax impression method: The shallow impressions of the biting surfaces of anterior teeth were produced as described by hand tracing from wax impression method. A small quantity of silver amalgam powder (DPI alloy, Non-gamma 2 amalgam) mixed in small quantities of surgical spirit was sequentially added to the individual tooth impressions using a thin hairbrush (Camlin hairbrush size 0) and was allowed to evaporate for approximately 5 minutes. The residual metal powder records the size, shape, and anatomical position of the biting surfaces of anterior teeth. A radiographic image was produced on an intraoral dental X- ray film (Kodak E speed no.2) using 70 kVp, 10 mA, dental X-ray machine by allowing the central ray directed at 90 0 to the wax sheet surface. After the film was processed, the bite marks showed up as white teeth marks in a dark black background. The radiographic image was traced on a transparent sheet and laterality markers were added. The overlay was then saved as radiographic overlay [Figure 3].
- Xerographic-based method: The upper and lower study casts were placed on the glass plate of the photocopy machine with the incisal edges down. The accuracy of a standard office photocopier was determined and calibrated to produce 100% images.  A slight weight was set on the top of the casts to ensure maximum incisal edge contact with the glass. The ABFO No. 2 scale (Lightning Powder Co. Inc., Order Catalog No. 6-3875, Lightning Powder Co., Inc. 13386 International Parkway, Jacksonville, Florida 32218, U. S. A) was placed adjacent to left side of dental casts as a laterality marker. This arrangement was covered with a white cloth to prevent any escape of ambient light. It was then photocopied on an A4 sized plain white paper. A transparent sheet was then overlaid on the photocopy image of the casts and the outlines of the incisal edges were traced. The laterality marker and serial number was added. The overlay was then scanned and the image was saved as xerography overlay [Figure 4].
Scanning the overlay
The transparent sheet with the overlay was laid upside down on the glass platen of the 2D scanner (Hewlett Packard Scan jet 3030 C flat bedded scanner with color laser printer). An ABFO No.2 bite mark standard reference scale was placed upside down over this sheet on the left side of the overlay and both were covered with a white sheet and scanned. The image of the overlay was acquired using the TWAIN interface of the Photoshop application and was saved as a JPEG file as per the method used.
- 2D Computer based method: In this method windows system with Intel Pentium 4 (1.6 GHz) Windows 98, a hard drive with 128 MB of random access memory, 8 bit (256 colors) display adapter card, a CD-ROM drive and color monitor was used.
- Scanning subjects casts: The study casts were positioned on the 2D scanner platen such that as many of the incisal edges contact or are close to the platen as possible. An ABFO No. 2 scale was placed at the lower left side of the casts to establish left laterality and ensure life sized reproduction. A black cloth was placed over the arrangement to eliminate ambient light. The scanning procedure was carried in a similar manner as described earlier except that the color photograph was chosen in the out put type field in the setting menu.
- Rotation of the image: The saved image was imported into Photoshop (Adobe Photoshop 6 software) and was rotated to make the edge of the scale parallel to the x-axis of the computer.
The hollow volume overlay was produced by the following steps:
- Selection of biting edges: The biting edge of teeth was selected using magic wand tool. The tolerance was set to 20 and the anti-aliased box was unchecked. The magic wand tool selects an area of similar pixel tone [Figure 5]. Once the initial selections in all the six teeth are done, the selection was smoothed.
- Construction of overlay: A new layer was created by clicking Layer>New. Selected edges were outlined by clicking Edit>Stroke; Stroke Width>1. Hollow volume overlay of the biting edges of anterior teeth appears [Figure 6].
- Adding the scale: The background layer was selected and with rectangular marquee tool the entire scale was selected. The selected area was stroked. Then the scale was copied and pasted into new layer. The required text viz, case number, arch, date etc was added to this image. This image was saved as a JPEG file in the same folder as computer-generated overlay.
Metric analysis on a computer: Metric analysis was carried out using NIH ImageJ software. The saved JPEG file was opened in ImageJ followed by clicking Image Menu>Type>8 Bit. A straight-line tool was chosen and in the image a straight line was drawn between two incremental lines representing 1 cm on the scale. This was followed by choosing Analyze menu>Measure. Then the required measurements were set. In the set measurement dialog box, area, centroid and limit to threshold were checked followed by entering decimal places>2.
The spatial scale of the active image was defined by clicking analyze menu>set scale>78 in distance in pixels>10 in known distance>mm in unit of measurement. The threshold was set by adjusting the top slide bar of the threshold dialogue box to the left to a value of zero and the bottom one to the right to 150.
Measurement of area of biting surfaces: With a Magic Wand Tool, each tooth mark area was selected, analyzed, measured in square mm, and saved as area results [Figure 7].
Measurement of tooth rotation
- The foreground color of the overlay was set to black using color picker tool and the cross hair (Mark and Count) tool was positioned in the tooth mark. The cross hair tool was moved over the tooth mark till the pixel coordinates in the message area as near as possible and clicked to mark the centroid. The centroid is the center point of the tooth mark. The centroid is marked for remaining biting edges in a similar manner.
- If the central incisors did not contact, a rectangular area was selected using a rectangular selection tool, with the horizontal guidelines positioned adjacent or above the top most and bottom most pixels of the central incisor and the vertical guides contacting the mesial contact points of the central incisors. If the central incisors overlapped, then the vertical guidelines of the rectangle should pass through the centroid of the central incisors. The rectangular area was measured and using cross hair tool and the centroid was marked. This point was considered as the mid point of the arch.
- A rectangular box of the image of upper and lower arch was created using rectangular selection tool and the image was oriented, so that the distal most pixel of the right and left canine teeth are in the same plane and contacts the horizontal line of the rectangle. The horizontal line intersecting the distal most pixels of the canines was termed as base line [2 in [Figure 8]]. Angle tool was selected with one arm bisecting the space between the central incisors through the mid point of the arch and the other arm coinciding with the baseline. A line perpendicular to the base line bisecting the space between the central incisors through the mid point of the arch was termed as Zero degree reference line (ZDRL) [1 in [Figure 8]]. The angle between the zero degree reference line to the base line should be 90 degree.
|Figure 8: Measurement of tooth rotation: (1) Zero degree reference line (ZDRL). (2) Base line|
Click here to view
- An angle tool was selected and relative rotation of each tooth was measured in degrees by the angle [1 in [Figure 9]] formed from the zero degree reference line (ZDRL) to a line [3 in [Figure 9]] connecting the centroid and mesial contact points of each biting surface or by recording the angle formed from a line parallel to the zero degree reference line [2 in [Figure 9]] to a line connecting the centroid and the mesial contact point.
|Figure 9: Measurement of tooth rotation: (1) Angle formed between zero degree reference line to a line connecting the centroid and mesial contact points of biting surface. (2) Angle formed from a line parallel to the ZDRL to a line connecting the centroid and the mesial contact point. (3) A line connecting the centroid and mesial contact points of each biting surface|
Click here to view
The overall results obtained were recorded for all the five different methods with respect to area and rotation of biting surfaces of twelve anterior teeth.
| Results|| |
For statistical analysis, the mean values obtained from measurements of area of biting edge and the relative rotation of six upper and six lower anterior teeth of 60 samples were calculated. Mean of each parameter, standard deviation, and standard error for all the 30 samples in the bite mark were calculated for all the five methods and are enumerated in the [Table 1] and [Table 2].
|Table 1: Mean, standard deviation, and standard error of area for 30 samples in the bite mark based on different methods|
Click here to view
|Table 2: Mean, standard deviation, and standard error of rotation for 30 samples in the bite mark based on different methods|
Click here to view
Data for area and rotation derived from the various production techniques were analyzed statistically using the MANOVA or multivariate analysis of variance. Results indicate the difference between the overlay techniques is statistically significant for area (P=0.045) and rotation (P=0.006). This indicates that there are differences between the methods in terms of analysis of all the physical characteristics when compared together. Variation with respect to rotational data between various techniques appears to be much less than for area.
The Mahanalobis distance between each test method and the computer-based method was measured to reach conclusions about relative accuracy. Two results were derived for each overlay production using this test. The first is the actual measure (unit value) of the distance from the test measurement to the mean of the gold standard data [Table 3].The second ranks the various overlay production methods in terms of decreasing accuracy (increasing Mahalanobis distance values). The accuracy of each method is determined by the amount of separation from gold standard [Table 4].
|Table 3: Mahalanobis distances for each overlay method compared to computer-based method|
Click here to view
|Table 4: Mahalanobis distance ranked in decreasing order of accuracy for area and rotation|
Click here to view
| Discussion|| |
Comparison techniques used in bite mark analysis are many and varied; the choice of technique depends mainly on the technology resource in forensic laboratories. Until recently, little research has been carried out to compare different bite mark analysis methods. Although various methods are used to produce bite mark overlays, inherent subjective error incorporated to the overlays may make it difficult to reach conclusions with a high degree of confidence in court proceedings. Hence to overcome these problems, computer-based techniques using scanner and adobe photoshop software are used to produce bite mark comparison overlays.
The present study was undertaken to evaluate the reliability and accuracy of five commonly used methods of human bite mark overlays, i.e., xerographic method, hand tracing from wax method, hand tracing from study casts, radiopaque wax method, and computer-based method. The overlay techniques were evaluated with respect to how accurately each reproduced the shape, size, and rotation of the upper and lower anterior teeth. The 2D computer-based method produced digital images of dental study casts, which were found to be very accurate. This method was taken as "gold standard" and other methods were compared to it to determine their accuracy. The area of the biting edge and relative rotation of each upper and lower front tooth were recorded and compared. On statistical analysis, the xerographic overlay production method was found to be the most accurate method for tooth area and rotation amongst the four methods. The results indicate that there is considerable variation in the determination of incisal edge area among the four overlay production methods. However, variation with respect to rotational data between the methods appears to be much less than for the area.
Area: The statistical analysis of the results obtained in this study shows that the xerographic method was the best among the four other different methods to measure the area of biting edges followed by hand tracing from wax method, hand tracing from study casts, radiopaque wax method in decreasing order of accuracy. This finding is not in agreement with the similar study conducted previously  where in authors have concluded that the radiographic overlay production is the most accurate method in determining the area of biting surfaces of twelve anterior teeth. This variation may be attributed to the fact that the area increases with the depth of the bite and the indentations of teeth created on a wax sheet from the study models depends on the amount of pressure applied. If the pressure is more, the area also increases and hence area measurement in radiopaque wax method and hand traced from wax method can vary among various examiners. Because of this many authors are of the opinion that the area of the biting surface is a highly unreliable physical characteristic to be considered in the metric analysis of bite marks.
The other reason for variation in determining the accuracy of area of the biting edges could be due to the subjective error introduced while tracing the radiographic image before scanning. Whereas in the previous study by Sweet et al.  the radiographic image was directly scanned thereby reducing possible subjective error. One more reason could be the magnification/distortion of radiographic image itself could have contributed to the variation.
Rotation: In this study, xerographic method of constructing the overlay to determine the tooth rotation ranks as the most accurate method. This finding is in agreement with the similar study conducted previously. This is followed by radiopaque wax method, hand tracing from wax method, and hand tracing from study cast.
The interpretation of tooth width, centroid position and zero degree reference line for determining the relative rotation of each tooth mark is relatively accurate and do not affect the determination of rotational values. The rotational data obtained in this study suggests that not much difference is seen between the four methods of overlay production and the difference between each overlay technique is statistically insignificant.
The use of photocopier calibrated to produce 100% images appears to be superior to other methods to record tooth rotation and area. However, the photocopier generated overlay technique does not replace sophisticated techniques that involve computer-generated overlays but provides a fast, effective method for comparisons that a lay jury can easily comprehend. The added advantage is that it can be made without the use of expensive equipment and could hence prove useful for preliminary screening purposes.
Recently developed new software package, "Dental Print" (2004, University of Granada, Department of Forensic Medicine and Forensic Odontology, Granada, Spain) will generate comparison overlays from 3D images of the suspect's dental cast. This software allows users to accurately and objectively select the biting edges of interest from the suspect's teeth when compared to 2D images. The procedure for generating comparison overlays is entirely automatic and it is impossible for third parties to manipulate or alter the 3D images. This dental print software is an important step forward in Forensic Sciences for bite mark analysis.  Irrespective of 2D or 3D methodology used, forensic odontologists must ensure that the techniques they use are supported by scientific evidence and the error rates should always be explained to the court to promote justice.
| Conclusion|| |
Not withstanding the limitations of each overlay production method identified in this study, it is recommended that hand tracing methods which depend on subjective input by the odontologists be discontinued as there is lot of scope for manipulation and observer bias. It is essential to develop a solid scientific foundation for bite mark analysis and further research in the comparison process is needed to enhance the reliability of bite mark analysis.
| References|| |
|1.||Clark DH. Practical Forensic Odontology. Oxford: Wright Publishers; 1992. |
|2.||Barsley RE. Forensic and Legal issues in Oral Diagnosis. Dent Clin North Am 1993;37:133-56. |
|3.||Rawson RD, Ommen RK, Kinard G, Johnson J, Fantis A. Statistical evidence for the individuality of the human dentition. J Forensic Sci 1984;29:245-53. |
|4.||Reider SF, Raymond RD, Barton GM, Ngoc N. Computer comparison of bite mark patterns in identical twins. J Am Dent Assoc 1982;105:449-51. |
|5.||Cameron J, Sims BG. Bite marks in Forensic Dentistry. Cameron J, Sims BG, editors. Edinburg: Churchill Livingstone; 1974. |
|6.||Tendeschi CG, Eckert WG, Tendeschi LG. Forensic Medicine: Physical Trauma. Vol 2. Philadelphia: WB Saunders Company; 1977. |
|7.||McCullough DC. Rapid comparison of bite marks by xerography. Am J Forensic Med Pathol 1983;4:355-8. |
|8.||Kouble RF, Craig GT. A comparison between direct and indirect methods for human bite mark overlays. J Forensic Sci 2004;49:111-8. |
|9.||Dailey JC. A practical technique for the fabrication of transparent bite mark analysis. J Forensic Sci 1991;36:565-70. |
|10.||Naru AS, Dykes E. The use of digital imaging technique. Sci Justice 1996;36:47-50. |
|11.||Bowers CM, Johansen RJ. Digital analysis of bite marks and human identification. Dent Clin North Am 2001;45:327-44. |
|12.||Sweet D, Parhar M, Wood RE. Computer based production of bite mark overlays. J Forensic Sci 1998;43:1050-5. |
|13.||McNamee AH, Sweet D, Pretty I. A comparative reliability analysis of computer-generated bite mark overlays. J Forensic Sci 2005;50:400-5. |
|14.||Sweet D, Bowers CM. Accuracy of bite mark overlays: A comparison of five common methods to produce exemplars from a suspect dentition. J Forensic Sci 1998;43:362-7. |
|15.||Pretty IA, Sweet D. Adherence of forensic odontologists to the ABFO bite mark guidelines for suspect evidence collection. J Forensic Sci 2001;46:1152-8. |
|16.||Martin-de las Heras S, Valenzuela A, Ogayar C, Valverde AJ, Torres JC. Computer based production of comparison overlays from 3D-scanned dental casts for bite mark analysis. J Forensic Sci 2005;50:127-33. |
Department of Oral Medicine and Radiology, Vydehi Institute of Dental Science and Research Center, Bangalore, Karnataka
Source of Support: None, Conflict of Interest: None
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9]
[Table 1], [Table 2], [Table 3], [Table 4]