Nodule-Aligned Latent Space Learning with LLM-Driven Multimodal Diffusion for Lung Nodule Progression Prediction
This addresses the challenge of early lung cancer diagnosis for patients with lung nodules, representing a domain-specific incremental improvement.
The paper tackled the problem of predicting lung nodule progression for early lung cancer diagnosis by proposing a framework that generates 1-year follow-up CT images from baseline scans and EHR data, achieving an AUROC of 0.805 and AUPRC of 0.346 for malignancy prediction, which outperforms baseline and state-of-the-art methods.
Early diagnosis of lung cancer is challenging due to biological uncertainty and the limited understanding of the biological mechanisms driving nodule progression. To address this, we propose Nodule-Aligned Multimodal (Latent) Diffusion (NAMD), a novel framework that predicts lung nodule progression by generating 1-year follow-up nodule computed tomography images with baseline scans and the patient's and nodule's Electronic Health Record (EHR). NAMD introduces a nodule-aligned latent space, where distances between latents directly correspond to changes in nodule attributes, and utilizes an LLM-driven control mechanism to condition the diffusion backbone on patient data. On the National Lung Screening Trial (NLST) dataset, our method synthesizes follow-up nodule images that achieve an AUROC of 0.805 and an AUPRC of 0.346 for lung nodule malignancy prediction, significantly outperforming both baseline scans and state-of-the-art synthesis methods, while closely approaching the performance of real follow-up scans (AUROC: 0.819, AUPRC: 0.393). These results demonstrate that NAMD captures clinically relevant features of lung nodule progression, facilitating earlier and more accurate diagnosis.