U.S. flag

An official website of the United States government, Department of Justice.

Testing Repeatability and Error of Coordinate Landmark Data Acquired From Crania

NCJ Number
Journal of Forensic Sciences Volume: 53 Issue: 4 Dated: July 2008 Pages: 782-785
Date Published
July 2008
4 pages

This pilot study evaluated the repeatability and error associated with the collection of coordinate cranial landmark data by means of direct digitization from dry skulls.


Unlike type one and type two cranial landmarks, which were found to be biologically informative in all directions, type three landmarks had significant digitizing error, leading the authors to caution researchers about relying on type three landmarks when selecting landmarks for coordinate data evaluation. Type one landmarks are defined as discrete juxtapositions of tissues, such as at the intersection of three sutures, for example, dacryon and asterion. Type two landmarks are curvature maxima or other local morphogenetic processes, usually with a biomechanical implication, such as a muscle attachment site. Ectoconchion and prosthion would be examples of type two landmarks. Type three landmarks, on the other hand, are extreme points, such as the endpoints of maximum cranial length and breadth. Type three landmarks are fairly accurate when instrumentally derived as linear distance in traditional morphometrics; however, these results are highly variable when attempting to archive their exact anatomical location. Moreover, type three landmarks are associated with a significant degree of error, both between and within observers. The authors recommend using only type one and type two landmarks as being biologically significant. The study collected the coordinates for 19 standard homolgous cranial landmarks using a Microscribe 3DX and G2X digitizer and the software Three-Skull written by Steve Ousley. The landmarks were selected to include standard type one and type two landmarks, as well as standard caliper-derived type three anatomical points. Three skulls were randomly selected from teaching specimens housed at the C.A. Pound Human Identification Laboratory. Each skull underwent three separate digitization sessions by two observers for a total of six digitizations for each skull. 4 tables, 3 figures, and 17 references

Date Published: July 1, 2008