Machine learning approaches for sex estimation using cranial measurements.

The aim of the present study is to apply support vector machines (SVM) and artificial neural network (ANN) as sex classifiers and to generate useful classification models for sex estimation based on cranial measurements. Besides, the performance of the generated sub-symbolic machine learning models is compared with models developed through logistic regression (LR). The study was carried out on computed tomography images of 393 Bulgarian adults (169 males and 224 females). The three-dimensional coordinates of 47 landmarks were acquired and used for calculation of the cranial measurements. A total of 64 measurements (linear distances, angles, triangle areas and heights) and 22 indices were calculated. Two datasets were assembled including the linear measurements only and all measurements and index, respectively. An additional third dataset comprising all possible interlandmark distances between the landmarks was constructed. Two machine learning algorithms-SVM and ANN and a traditional statistical analysis LR-were applied to generate models for sex estimation. In addition, two advanced attribute selection techniques (Weka BestFirst and Weka GeneticSearch) were used. The classification accuracy of the models was evaluated by means of 10 × 10-fold cross-validation procedure. All three methods achieved accuracy results higher than 95%. The best accuracy (96.1 ± 0.5%) was obtained by SVM and it was statistically significantly higher than the best results achieved by ANN and LR. SVM and ANN reached higher accuracy by training on the full datasets than the selection datasets, except for the sample described by the interlandmark distances, where the reduction of attributes by the GeneticSearch algorithm improved the accuracy.

Data mining for sex estimation based on cranial measurements.

The aim of the present study is to develop effective and understandable classification models for sex estimation and to identify the most dimorphic linear measurements in adult crania by means of data mining techniques. Furthermore, machine learning models and models developed through logistic regression analysis are compared in terms of performance. Computed tomography scans of 393 adult individuals were used in the study. A landmark-based approach was applied to collect the metric data. The three-dimensional coordinates of 47 landmarks were acquired and used for calculation of linear measurements. Two datasets of cranial measurements were assembled, including 37standard measurements and 1081 interlandmark distances, respectively. Three data mining algorithms were applied: the rule induction algorithms JRIP and Ridor, and the decision tree algorithm J48. Two advanced attribute selection methods (Weka BestFirst and Weka GeneticSearch) were also used. The best accuracy result (91.9 %) was achieved by a set of rules learnt by the JRIP algorithm from the dataset constructed by application of the GeneticSearch selection algorithm to the dataset of standard cranial measurements. The set consisted of five rules including seven cranial measurements. Its accuracy was even better than the classification rates achieved by the logistic regression models. Concerning the second dataset of nonstandard measurements, the best accuracy (88.3 %) was obtained by using classification models learnt by two algorithms - JRIP with a dataset preprocessed by the BestFirst selection algorithm and Ridor with preprocessing by the GeneticSearch selection algorithm. Our experiments show that for the two datasets mentioned above the rule-based models contain smaller sets of rules with shorter lists of measurements and achieve better classification accuracy results in comparison with decision tree-based models.

Sex estimation by Mastoid Triangle using 3D models.

Mastoid Triangle (MT) is a triangle constructed between the landmarks porion, mastoidale and asterion. The aim of the study is to evaluate the sex differences in the MT size in Bulgarian adults and to develop discriminant functions for sex estimation based on the MT sides and area. The study was carried out on 148 head CT scans. A sample of 53 males and 53 females was used for development of discriminant functions, and a test sample of 21 males and 21 females was applied for their validation. Using the software InVesalius©, 3D models of the skulls were segmented and exported in STL format. The 3D coordinates of the landmarks porion, asterion and mastoidale were collected using the software MeshLab©. The MT sides, area and angles were calculated. The sex differences were assessed by the independent t-test. Bilateral differences were evaluated using the paired t-test. Univariate and multivariate discriminant function analyses were applied. The results showed that the MT sides and area differed significantly between both sexes. Sex differences were also established for the angle at mastoidale. Bilateral differences were found in males for the distance porion-mastoidale, which was significantly greater on the right side. The MT dimensions showed sufficient discriminating power for sex estimation among Bulgarians (up to 89%), and the total MT area proved to be the best single sex discriminating trait. The test sample corroborated the usefulness of the MT in sex estimation demonstrating similar or higher overall accuracy rates.

Sex estimation by size and shape of foramen magnum based on CT imaging.

Foramen magnum (FM) has a well-protected position, which makes it of particular interest in forensic research. The aim of the study is to assess the sex differences in size and shape of FM, develop discriminant functions and logistic regression models based on the FM measurements, compare the accuracy results of the measurements obtained through different measuring approaches, and establish the most reliable variables for sex estimation in Bulgarian adults. Head CT scans of 140 Bulgarian adults were used in the study. The segmentation of the skulls was performed in the software InVesalius. The length, breadth, circumference, and area were measured based on the 3D coordinates of definite landmarks and semi-landmarks. The circumference and area were calculated regarding the foramen as a 2D and 3D structure. Two additional variables (λ2 and λ3) corresponding to the least square errors along the length and breadth directions at the fitting of the 3D coordinates to a plane were examined for their sex discriminating ability. The FM shape was classified based on the values of the FM index. The significance of the sex differences was assessed. Discriminate function analysis and binary logistic regression were conducted. Significant sex differences were established in the FM size and shape. The eigenvalue λ3 is the best discriminating parameter applying discriminant function analysis. The acceptance of FM as a 2D or 3D structure does not provide substantial information for its sex discrimination. The measurements of FM do not offer sufficiently high predicting rates for sex estimation in the Bulgarian population.

Evaluation of the effectiveness of CT-guided infiltration in the treatment of Arnold's neuralgia.

INTRODUCTION: The aim of this study was to evaluate the effectiveness of computed tomography (CT)-guided infiltration in the treatment of Arnold's neuralgia. METHODS: A retrospective study included 31 patients suffering from Arnold's neuralgia and having undergone a total of 45 CT-guided infiltrations of the greater occipital nerve (GON), in a proximal site (emergence of the GON, technique 1, n = 24) or in two proximal sites (emergence of the GON and at the site of the first bend of the GON drawn by the GON, technique 2, n = 21). Infiltration was considered to be effective when pain relief was equal to or greater than 50% for at least 1 month. RESULTS: There was no significant difference between the two techniques regarding immediate pain relief effect (53.3% for technique 1 vs. 60.5% for technique 2, p = 0.5), but technique 2 yielded better persistence of pain relief effect (p = 0.01), leading to a significantly higher percentage of effective infiltrations with technique 2 (p = 0.03). CONCLUSION: Infiltrations carried out in a single site yield results that are comparable to those previously published. Infiltrations in two sites provide significantly better results and should now be preferred to other single-site techniques in order to reduce the rate of failure or recurrence of Arnold's neuralgia.