Search

This study aimed to develop image-analysis-based classification models for distinguishing individuals younger and older than 30 using the medial clavicle. We extracted 2D images of the medial clavicle from multi-slice computed tomography (MSCT) scans from Clinical Hospital Center Split (n = 204). A sample was divided into a training (164 images) and testing (40 images) dataset. The images were loaded into the Orange Data Mining 3.32.0., and transformed into vectors using the pre-trained neural network Painters: A model trained to predict painters from artwork images. We conducted Principal Components Analysis (PCA) to visualize regularities within data and reduce data dimensionality in classification. We employed three classifiers that provided >80% accuracy: Support Vector Machine (SVM), Logistic Regression (LR), and Neutral Network Identity SGD (NNI-SGD). We used 5-fold cross-validation (CV) to obtain optimal variables and performances and validated data on the independent test set, with a standard posterior probabilities (pp) threshold of 0.5 and 0.95. The explainability of the model was accessed visually by analyzing clusters and incorrectly classified images using anthropology field knowledge. Based on the PCA, clavicles clustered into categories under 30 and 40 years, between 40 and 55 years, and over 80 years. The overall accuracy with standard pp ranged from 82.5% to 92.5% for CV and 82.5% to 92.5% for the test set. The posterior probability of 0.95 provided classification accuracy up to 100% but with a lower proportion of images that could be classified. The study showed that image analysis based on a pre-trained deep neural network could contribute to distinguishing clavicles of individuals younger and older than 30., Competing Interests: The authors have declared that no competing interests exist., (Copyright: © 2024 Ivković et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.)

This study investigated the stroke and slope characteristics in left-handed and right-handed handwriting. Stroke (letters t, f, đ, and H) and slope (letters t, f, l, d, and g) directions were analyzed on in-house samples (n = 64), revealing statistically significant differences (p ≤ 0.05) between the groups. Right-handers predominantly exhibited left-to-right strokes (98%-100%), while left-handers showed greater variability. Although statistically significant for most letters analyzed, slope direction did not demonstrate consistent patterns. A logistic regression model was developed and validated on the same sample to classify handedness based on the averaged strokes of the letter "t." The model was further tested on samples (n = 252) from a publicly available handwriting database. If the model classified the sample as produced by left hand, it was correct in 100% of cases. In contrast, when the model classified writing as right-handed, it was correct in 73%-97% of cases, depending on the validation sample. The model classified writing as of left-handed origin if more than 36% of the letters "t" had a stroke from right to left, while otherwise, writing was classified as of right-handed origin. The developed method showed great potential for classifying the handedness of the author of disputed handwriting, thus eliminating individuals as text authors or narrowing down the pool of potential authors., (© 2024 American Academy of Forensic Sciences.)

This study investigates a series of antemortem injuries in the skeletal remains of a 36-45-year-old male from the medieval site of Rižinice. It presents the injuries comprising a fractured sternum, spinal hyperflexion, multiple rib fractures, and a non-union fracture to the olecranon process of the left ulna and a patella cubiti , a rare anomaly, which until now has not been reported in a paleopathological context. The research aims to uncover the causes, timing, and effects of these traumata, providing insight into the challenges of life in the mid-9th to mid-10th century., Competing Interests: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (© 2024 The Authors.)

Aim: To determine the incidence of metopism in the modern and archaeological Croatian population., Methods: A total of 800 specimens (454 modern multi-slice computed tomography [MSCT] scans and 346 dry archaeological skulls) were visually examined for metopic suture presence. The metopic suture was deemed complete when aligned nasion to bregma., Results: In the overall sample, the metopic suture was observed in 36 of 800 subjects (4.5%): 19 of 424 (4.5%) men and 17 of 370 (4.6%) women. A significant difference was not observed between modern and archaeological samples (χ2=3.219, P=0.359) or between the sexes (χ2=0.006, P=0.939). The frequency of metopism varied from 3.5% in the modern population to 7.04% in the samples from the Roman period., Conclusion: There are no visible secular changes on metopic suture in the Croatian population through time. Some variations can be the result of differences in sample size in different time periods.

The aim of this study was to test the sexual dimorphism of orbital measurements in the Croatian population using multi-slice computed tomography (MSCT) images. We have retrospectively taken 414 head CT scans of adults from Croatian clinical hospitals in Split and Zagreb (214 males and 200 females) with slice thickness < 1 mm and no pathological or traumatic changes that could affect the measurements. DICOM files were imported into Stratovan Checkpoint Software and viewed in 2D and 3D using semi-transparent 3D volume rendering. Eight standard measurements were calculated based on twelve orbital landmarks (six paired). Principal component analysis (PCA) was used to explore sexual and regional differences, and linear discriminant analysis was used to develop sex classification models. The PCA showed separation based on sex and region, and additional analysis demonstrated that females and males in Split and Zagreb differed in four orbital measurements (P ≤ 0.001). Only those measurements that did not show regional differences were further analyzed, and all showed statistically significant sexual dimorphism. The accuracy of univariate functions for sex estimation ranged from 53.43 to 71.88%, and for multivariate function, the accuracy was 73.45%. The orbital measurements of the Croatian population showed restricted forensic significance for sex classification. On the other hand, we have shown that they can have a potential for exploring the inter- and intra-population differences., (© 2022. The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.)

Background: Blood group phenotypes have been associated with COVID-19 susceptibility and severity. This study aimed to examine ABO/Rh blood group distribution in COVID-19-related deaths considering demographics and pathological conditions., Methods: We conducted a retrospective study at the University Hospital Center Split, Croatia, that included 245 COVID-19 positive individuals that died from April 8, 2020, to January 25, 2021. We extracted data on their blood groups, demographics, and pre-existing comorbidities and compared findings with general population data from blood group donations (n = 101,357) and non-COVID-19 deaths from 2019 (n = 4968)., Results: The proportion of dead males was significantly higher than in non-COVID-19 cases (63.7% vs. 48.9%, P < 0.001), while the proportion of older individuals did not differ. The prevailing pre-existing diseases were hypertension (59.6%), diabetes (37.1%), heart failure (28.8%), digestive disorder (26.5%), and solid tumor (21.6%). The ABO distribution in the deceased and donors' group showed significant differences, with the higher prevalence of A/AB group and lower prevalence of 0, but with individual differences significant only for AB and non-AB groups. There was a reduced proportion of females within the deceased with group 0 (P = 0.014) and a higher proportion of AB individuals with coronary heart disease (P = 0.024)., Conclusion: The study confirmed a higher risk of death in males. The lower proportion of type 0 in deceased individuals was greater in females, implying that group 0 is not necessarily an independent protective factor. Coronary heart disease was identified as a potential risk factor for AB individuals., Competing Interests: Declaration of Competing Interest Authors declare no conflict of interest., (Copyright © 2022 Elsevier Ltd. All rights reserved.)

This study examined if the cranial measurements from Data Collection Procedures for Forensic Skeletal Material 2.0 are repeatable when measured in dry bones and MSCT images and if the virtual measurements correspond to the physical ones. The sample included 33 dry crania imaged by MSCT. Two observers measured dry bones, two placed landmarks on 2D and 3D MSCT reconstructions, and one conducted measurements/landmarking on both media. One of the observers for each media repeated the measurements. Technical and relative technical error of measurement (TEM and rTEM) and percentage differences were calculated to examine the repeatability of measurements and compare measuring modalities. Intraobserver rTEM was above 1.5% for six bone measurements: FOB, ZOB, OBB, NLH, DKB, MDH (1.51%-4.87%) and for seven MSCT measurements: OBH, FOB, OBB, MDH, NLB, ZOB, DKB (1.57%-5.55%). The interobserver rTEM was above the acceptable level (>2%) for 11 measurements: PAC, NLH, OBB, EKB, MAL, FOB, NLB, OBH, ZOB, DKB, and MDH (2.01%-9.34%). The percentage differences were not systematically larger for measurements taken by the same user on both modalities than those obtained by different users on the same modality. When physical and MSCT measurements were tested on sex classification standards, the proportion of crania classified as male or female did not significantly differ (p > 0.05). The study showed that physical and virtual cranial measurements could be interchangeable for developing or applying sex estimation standards. However, clarifications and adaptations are necessary for measurements of mastoid, nasal, and orbital regions that did not meet the standard criteria., (© 2022 American Academy of Forensic Sciences.)

Aim: To analyze an additional set of ˝Y-chromosome genetic markers to acquire a more detailed insight into the diversity of the Croatian population., Methods: A total of 518 Yfiler Plus profiles were genotyped. Allele frequencies, haplotype frequencies, and haplotype diversity were calculated by using the STRAF software v. 2.0.4. Genetic distances were quantified by Rst with AMOVA online tool from the YHRD. The evolutionary history was inferred with the neighbor-joining method of phylogenetic tree construction in the MEGAX software. Whit Athey's Haplogroup Predictor v. 5 was used for additional comparison with regional and other European populations., Results: A total of 507 haplotypes were used for genetic STR analysis. An interpopulation study on 17 Y-STR markers showed the lowest genetic diversity between the Croatian and Bosnian-Herzegovinian populations and the highest between the Croatian and Irish populations. Additional interpopulation comparison with the original 27 Y-STR markers (for the population with available data) was also performed. A total of 518 haplotypes were used in the determination of haplogroup diversity. Haplogroup I with its sublineage I2a expressed the highest prevalence. The second most prevalent haplogroup was R, with its major sublineage R1a, except for the subpopulation of Hvar, where E1b1b was the second most prevalent haplogroup. Rare haplogroups also confirmed in this study were L, T, and Q. G1 was detected for the first time in the Croatian population., Conclusion: We obtained a new insight into the differences between examined subpopulations of Croatia and their possible (dis)similarities with neighboring and distant populations.

The adjusted binary classification (ABC) approach was proposed to assure that the binary classification model reaches a particular accuracy level. The present study evaluated the ABC for osteometric sex classification using multiple machine learning (ML) techniques: linear discriminant analysis (LDA), boosted generalized linear model (GLMB), support vector machine (SVM), and logistic regression (LR). We used 13 femoral measurements of 300 individuals from a modern Turkish population sample and split data into two sets: training (n = 240) and testing (n = 60). Then, the five best-performing measurements were selected for training univariate models, while pools of these variables were used for the multivariable models. ML classifier type did not affect the performance of unadjusted models. The accuracy of univariate models was 82−87%, while that of multivariate models was 89−90%. After applying ABC to the crossvalidation set, the accuracy and the positive and negative predictive values for uni- and multivariate models were ≥95%. Sex could be estimated for 28−75% of individuals using univariate models but with an obvious sexing bias, likely caused by different degrees of sexual dimorphism and between-group overlap. However, using multivariate models, we minimized the bias and properly classified 81−87% of individuals. A similar performance was also noted in the testing sample (except for FEB), with accuracies of 96−100%, and a proportion of classified individuals between 30% and 82% in univariate models, and between 90% and 91% in multivariate models. When considering different training sample sizes, we demonstrated that LR was the most sensitive with limited sample sizes (n < 150), while GLMB was the most stable classifier.