Nodular thyroid disease constitutes a prevalent thyroid condition, leading the incidence rate among thyroid disorders. It is particularly noteworthy that coal miners, exposed to coal dust over extended periods, face a significantly heightened risk of nodular thyroid disease due to their occupational exposure. To explore this issue further, this study enrolled 1,708 coal miners who underwent physical examinations at the Huaibei Energy Occupational Disease Prevention and Treatment Hospital in Anhui Province in April 2021. We meticulously gathered comprehensive clinical data, encompassing general information, laboratory test outcomes, and imaging examination results. Our research introduces a novel Non-linear Inverse Nearest Manifold Projection (NKLPP) model to evaluate the risk of nodular thyroid disease in coal miners. The model leverages advanced nonlinear mapping techniques to project high-dimensional data into a low-dimensional manifold space, thereby capturing the data's intrinsic structure and patterns to facilitate more precise identification of disease risk factors. Moreover, we conducted performance comparisons with several established risk assessment models. The findings demonstrate that the NKLPP model achieves exceptional performance in both F1 score and Area Under the Precision-Recall Curve (AP) metrics, significantly surpassing other models and offering substantial clinical utility.