Back to today’s briefing

Story perspectives

Deep Neural Networks Revolutionize Sternum Sex Estimation in Turkey

9/29/2025

46 10

Summary

Story summary

First to apply deep neural networks (DNN) to sternum CT sex estimation in Turkey.
It compares DNN with k-Nearest Neighbors (KNN) and Linear Discriminant Analysis (LDA).
DNN achieves AUC 0.937; LDA shows highest performance, not statistically different.
Limitations include a small dataset and lack of external validation.
Future work seeks diverse data and hyperparameter tuning; DNN offers automation for forensics.

1 of 1

Story summary

First to apply deep neural networks (DNN) to sternum CT sex estimation in Turkey.
It compares DNN with k-Nearest Neighbors (KNN) and Linear Discriminant Analysis (LDA).
DNN achieves AUC 0.937; LDA shows highest performance, not statistically different.
Limitations include a small dataset and lack of external validation.
Future work seeks diverse data and hyperparameter tuning; DNN offers automation for forensics.