SexEst: An open access web application for metric skeletal sex estimation

Abstract

Skeletal sex estimation is an essential step in any osteoarcheological study; hence, several metric and morphological methods have been developed for this purpose, employing different skeletal elements. This paper has a dual purpose: (1) test the performance of several machine learning classification models for skeletal sex estimation using worldwide samples of cranial and postcranial measurements and (2) present a free web application for the implementation of the models that exhibit the highest accuracy so that the sex of unknown skeletons can be straightforwardly estimated. Regarding the first objective, using the Goldman database of postcranial metrics and the William W. Howells craniometric database, machine learning classification models were constructed for sex prediction. The models were optimized with respect to their hyperparameters and cross‐validated reaching accuracies ranging from 80.8%–89.5% for the postcranial data and 81.2%–87.7% for the cranial data. The models offering the highest rates of correct sex classification (Extreme Gradient Boosting, Light Gradient Boosting, and Linear Discriminant Analysis) were then selected to construct an open access and open source web application, SexEst, for predicting the sex of unknown skeletons.

Evaluation of the (hu)MANid program for sex and ancestry estimation in a diverse, contemporary CT scan‐based sample

“…Many times, the mandible can survive conditions that other bones in the human body cannot due to its robust nature [ 21 ]. There has been a movement toward developing easy‐to‐use, validated software‐based approaches for forensic skeletal analysis (e.g., FORDISC [ 23 ], CRANID [ 36 ]; SexEst [ 37 ]), but these software have tended to focus on the cranium or post‐cranial elements. To improve ease of evaluating a forensic profile from mandibular material, Berg and Kenyhercz [ 38 ] developed the (hu)MANid program, aimed at allowing sex and ancestry estimation using only mandibular features.…”

Testing the accuracy of the SexEst software for sex estimation in a modern Greek sample

“…However, most available software cannot handle missing values and/or uses population‐specific reference standards. This work aimed to examine the reliability of the open‐access SexEst software (Constantinou & Nikita, 2022), which has been developed based on a global sample of mostly preindustrial skeletal reference material, on a modern Greek population. Note that despite the global nature of the SexEst training dataset, its reference database does not include any Greek samples.…”

“…Correspondingly, the DSP2 software (Bru ˚žek et al, 2017) focuses on pelvic traits, while the SexEst software employs cranial and postcranial measurements collected from individuals with diverse ancestry (Constantinou & Nikita, 2022).…”

“…It additionally estimates sex based on postcranial measurements, but the reference sample consists of American Black and American White individuals (Manthey & Jantz, 2020). Correspondingly, the DSP2 software (Brůžek et al, 2017) focuses on pelvic traits, while the SexEst software employs cranial and postcranial measurements collected from individuals with diverse ancestry (Constantinou & Nikita, 2022).…”

“…Instead, in cases of incomplete datasets, the users must select one of the LDA models that predict sex with fewer variables. For a detailed description of the machine learning procedure and the overall development of the SexEst software, please see Constantinou and Nikita (2022).…”