Occupational dust exposure and thyroid nodule screening: A predictive modeling approach for early detection in coal mining communities
Feng Zhao, Kanghui Wu, Hongzhen ZhangBackground
This study aimed to identify risk factors and develop an advanced risk assessment model to evaluate the risk of nodular thyroid disease in coal miners.
Methods
In April 2021, 1708 coal miners undergoing physical examinations at the Huainan Energy Occupational Disease Prevention and Treatment Hospital in Anhui Province were enrolled in this study. Comprehensive clinical data were collected, including general information, laboratory test results, and imaging examination findings. A novel Nonlinear Inverse Nearest Manifold Projection model was developed to assess the risk of nodular thyroid disease. This model employs advanced nonlinear mapping techniques to project high-dimensional data into a low-dimensional manifold space, capturing the intrinsic structure and patterns of the data to identify disease risk factors more accurately. The performance of the Nonlinear Inverse Nearest Manifold Projection model was compared with several established risk assessment models.
Results
The Nonlinear Inverse Nearest Manifold Projection model demonstrated exceptional performance, achieving high scores in both F1 score and the area under the precision-recall curve metrics. It significantly outperformed other risk assessment models, highlighting its superior capability in identifying nodular thyroid disease risk factors among coal miners.
Conclusions
The Nonlinear Inverse Nearest Manifold Projection model is a highly effective tool for assessing the risk of nodular thyroid disease in coal miners, offering substantial clinical utility.