Development of a Machine Learning Nomogram to Assess Non-Suicidal Self-Injury Risk in Depressed Adolescents: A Multicenter Analysis

1

A machine learning nomogram was developed to assess non-suicidal self-injury risk in adolescents with depression.
2

The study analyzed data from 2,343 adolescents with depression across 14 hospitals in China.
3

Eight key variables were identified as predictors of NSSI, including depression score and perceived family support.
4

The nomogram demonstrated good performance with AUC values of 0.754 in training and 0.748 in validation cohorts.
5

This tool aims to enhance individualized risk assessment and clinical management for adolescents at risk of NSSI.

Frontiers In Psychiatry

by Lan Hong, Jianuo Shi, Qianjin Lou, Ying Yao, Tianshu Peng, Zhen Xu, Jinwei Gai, Zhaoxuan Liu, Siyu Tong, Tiansheng Zheng, Dongwu Xu, Ke Zhao
July 17, 2026

1