Mediastinal Lymph Node Detection and Segmentation Using Deep Learning

TL;DR

This paper presents a modified UNet deep learning model with advanced upsampling techniques for improved mediastinal lymph node detection and segmentation in CT images, achieving high accuracy and outperforming existing methods.

Contribution

The study introduces a novel UNet modification using bilinear interpolation and TGV-based upsampling to enhance lymph node segmentation in challenging CT images.

Findings

Achieved 94.8% accuracy on combined datasets.

Outperformed state-of-the-art segmentation approaches.

Demonstrated robustness across multiple datasets.

Abstract

Automatic lymph node (LN) segmentation and detection for cancer staging are critical. In clinical practice, computed tomography (CT) and positron emission tomography (PET) imaging detect abnormal LNs. Despite its low contrast and variety in nodal size and form, LN segmentation remains a challenging task. Deep convolutional neural networks frequently segment items in medical photographs. Most state-of-the-art techniques destroy image's resolution through pooling and convolution. As a result, the models provide unsatisfactory results. Keeping the issues in mind, a well-established deep learning technique UNet was modified using bilinear interpolation and total generalized variation (TGV) based upsampling strategy to segment and detect mediastinal lymph nodes. The modified UNet maintains texture discontinuities, selects noisy areas, searches appropriate balance points through backpropagation, and recreates image resolution. Collecting CT image data from TCIA, 5-patients, and ELCAP public dataset, a dataset was prepared with the help of experienced medical experts. The UNet was trained using those datasets, and three different data combinations were utilized for testing. Utilizing the proposed approach, the model achieved 94.8% accuracy, 91.9% Jaccard, 94.1% recall, and 93.1% precision on COMBO_3. The performance was measured on different datasets and compared with state-of-the-art approaches. The UNet++ model with hybridized strategy performed better than others.