Deep Learning on Point Clouds for False Positive Reduction at Nodule Detection in Chest CT Scans

Abstract

This paper focuses on a novel approach for false-positive reduction (FPR) of nodule candidates in Computer-Aided Detection (CADe) systems following the suspicious lesions detection stage. Contrary to typical decisions in medical image analysis, the proposed approach considers input data not as a 2D or 3D image, but rather as a point cloud, and uses deep learning models for point clouds. We discovered that point cloud models require less memory and are faster both in training and inference compared to traditional CNN 3D, they achieve better performance and do not impose restrictions on the size of the input image, i.e. no restrictions on the size of the nodule candidate. We propose an algorithm for transforming 3D CT scan data to point cloud. In some cases, the volume of the nodule candidate can be much smaller than the surrounding context, for example, in the case of subpleural localization of the nodule. Therefore, we developed an algorithm for sampling points from a point cloud constructed from a 3D image of the candidate region. The algorithm is able to guarantee the capture of both context and candidate information as part of the point cloud of the nodule candidate. We designed and set up an experiment in creating a dataset from an open LIDC-IDRI database for a feature of the FPR task, and is herein described in detail. Data augmentation was applied both to avoid overfitting and as an upsampling method. Experiments were conducted with PointNet, PointNet++, and DGCNN. We show that the proposed approach outperforms baseline CNN 3D models and resulted in 85.98 FROC versus 77.26 FROC for baseline models. We compare our algorithm with published SOTA and demonstrate that even without significant modifications it works at the appropriate performance level on LUNA2016 and shows SOTA on LIDC-IDRI. Keywords: Image analysis • Computer vision • Computer detection • Computer-assisted • Lung cancer • Pulmonary nodule • False-positive reduction • Point cloud • Chest CT Supported by BOTKIN.AI, Skolkovo.

Evaluation of the prospects for using artificial intelligence technologies to analyze CT scans of the chest organs in order to identify signs of malignant neoplasms in the lungs

“…и 3000 исследований, дополнительно содержащих исследования с признаками COVID-19 (дополнительные 500 исследований к 2500 первого датасета). Описание архитектуры и методики обучения моделей приведено в работах [11,12].…”

Researches advanced in application of medical image analysis based on deep learning

“…Instead of following the thinking pattern of 2D or 3D information, Drokin, Ivan and Elena Ericheva proposes a point cloud transformation to reduce the memory consumption of the model and speed up the training and inference process. What's more, it gives no restriction on input image size [19].…”

“…Yan, Ke et al builds a one-staged framework to allow the model to be trained end-to-end, which improves the model's training and inference speed on the basis of 3D CNN model [9]. On the other hand, Drokin, Ivan and Elena Ericheva try to jump out of the thinking pattern of 2D or 3D view of data, builds a point cloud model for FPR task to improve the model's computational speed compared with the traditional 3D CNN [19]. Although this work requires less memory and achieves better performance, it is designed for the second phase of the whole nodule detection task.…”