Deep Convolutional Architecture for Block-Based Classification of Small Pulmonary Nodules

Effect of CAD on Radiologists' Detection of Lung Nodules on Thoracic CT Scans: Analysis of an Observer Performance Study by Nodule Size

Accurate pulmonary nodule detection is a crucial step in lung cancer screening. Computer-aided detection (CAD) systems are not routinely used by radiologists for pulmonary nodule detection in clinical practice despite their potential benefits. Maximum intensity projection (MIP) images improve the detection of pulmonary nodules in radiological evaluation with computed tomography (CT) scans. Inspired by the clinical methodology of radiologists, we aim to explore the feasibility of applying MIP images to improve the effectiveness of automatic lung nodule detection using convolutional neural networks (CNNs). We propose a CNN-based approach that takes MIP images of different slab thicknesses (5 mm, 10 mm, 15 mm) and 1 mm axial section slices as input. Such an approach augments the two-dimensional (2-D) CT slice images with more representative spatial information that helps discriminate nodules from vessels through their morphologies. Our proposed method achieves sensitivity of 92.7% with 1 false positive per scan and sensitivity of 94.2% with 2 false positives per scan for lung nodule detection on 888 scans in the LIDC-IDRI dataset. The use of thick MIP images helps the detection of small pulmonary nodules (3 mm-10 mm) and results in fewer false positives. Experimental results show that utilizing MIP images can increase the sensitivity and lower the number of false positives, which demonstrates the effectiveness and significance of the proposed MIP-based CNNs framework for automatic pulmonary nodule detection in CT scans. The proposed method also shows the potential that CNNs could gain benefits for nodule detection by combining the clinical procedure.

Diminution index: A novel 3D feature for pulmonary nodule detection

Multiview two-dimensional (2D) convolutional neural networks (CNNs) and three-dimensional (3D) CNNs have been successfully used for analyzing volumetric data in many state-of-the-art medical imaging applications. We propose an alternative modular framework that analyzes volumetric data with an approach that is analogous to radiologists' interpretation, and apply the framework to reduce false positives that are generated in computer-aided detection (CADe) systems for pulmonary nodules in thoracic computed tomography (CT) scans. In our approach, a deep network consisting of 2D CNNs first processes slices individually. The features extracted in this stage are then passed to a recurrent neural network (RNN), thereby modeling consecutive slices as a sequence of temporal data and capturing the contextual information across all three dimensions in the volume of interest. Outputs of the RNN layer are weighed before the final fully connected layer, enabling the network to scale the importance of different slices within a volume of interest in an end-to-end training framework. We validated the proposed architecture on the false positive reduction track of the lung nodule analysis (LUNA) challenge for pulmonary nodule detection in chest CT scans, and obtained competitive results compared to 3D CNNs. Our results show that the proposed approach can encode the 3D information in volumetric data effectively by achieving a sensitivity >0.8 with just 1/8 false positives per scan. Our experimental results demonstrate the effectiveness of temporal analysis of volumetric images for the application of false positive reduction in chest CT scans and show that state-of-the-art 2D architectures from the literature can be directly applied to analyzing volumetric medical data. As newer and better 2D architectures are being developed at a much faster rate compared to 3D architectures, our approach makes it easy to obtain state-of-the-art performance on volumetric data using new 2D architectures.

Automatic detection of lung nodules in CT datasets based on stable 3D mass–spring models

Early diagnosis of lung cancer plays crucial role in the improvement of patients' chances of survival. Computer aided detection (CAD) system has been a groundbreaking step in the timely diagnosis and identification of potential nodules (lesions). CAD system starts detection process by extracting lung regions from CT scan images, this step narrows down the region for detection. Hence saving the time consumption and reducing false positives outside the lung regions that results in the improvement of specificity of system. Proper lung segmentation significantly increases the performance of CAD systems. Different algorithms are presented by various researchers to improve segmentation results. An intensity based approach is presented in this paper for the segmentation of parenchyma and the goal is to achieve reasonable segmentation results in less time. Algorithm used in this paper is based on the Intensity based thresholding which is the fastest method for image segmentation. Images used in this research to analyze algorithm's result are taken from Lung Image Database Consortium (LIDC). Twenty random cases were picked, each having different number of slices (128 to 300). Algorithm is implemented using MatlabR2014 and a system with processor of 2.6 GHz and RAM of 4 GB. Total time taken for a single case of 128 images was 6.3 seconds and hence with an average of 49 milli sec/slice.

To distinguish benign from malignant pulmonary nodules using CT images is critical for their precise diagnosis and treatment. A new Agile convolutional neural network (CNN) framework is proposed to conquer the challenges of a small-scale medical image database and the small size of the nodules, and it improves the performance of pulmonary nodule classification using CT images. A hybrid CNN of LeNet and AlexNet is constructed through combining the layer settings of LeNet and the parameter settings of AlexNet. A dataset with 743 CT image nodule samples is built up based on the 1018 CT scans of LIDC to train and evaluate the Agile CNN model. Through adjusting the parameters of the kernel size, learning rate, and other factors, the effect of these parameters on the performance of the CNN model is investigated, and an optimized setting of the CNN is obtained finally. After finely optimizing the settings of the CNN, the estimation accuracy and the area under the curve can reach 0.822 and 0.877, respectively. The accuracy of the CNN is significantly dependent on the kernel size, learning rate, training batch size, dropout, and weight initializations. The best performance is achieved when the kernel size is set to [Formula: see text], the learning rate is 0.005, the batch size is 32, and dropout and Gaussian initialization are used. This competitive performance demonstrates that our proposed CNN framework and the optimization strategy of the CNN parameters are suitable for pulmonary nodule classification characterized by small medical datasets and small targets. The classification model might help diagnose and treat pulmonary nodules effectively.

Computerized tomography (CT) is a clinically primary technique to differentiate benign-malignant pulmonary nodules for lung cancer diagnosis. Early classification of pulmonary nodules is essential to slow down the degenerative process and reduce mortality. The interactive paradigm assisted by neural networks is considered to be an effective means for early lung cancer screening in large populations. However, some inherent characteristics of pulmonary nodules in high-resolution CT images, e.g., diverse shapes and sparse distribution over the lung fields, have been inducing inaccurate results. On the other hand, most existing methods with neural networks are dissatisfactory from a lack of transparency. In order to overcome these obstacles, a united framework is proposed, including the classification and feature visualization stages, to learn distinctive features and provide visual results. Specifically, a bilateral scheme is employed to synchronously extract and aggregate global-local features in the classification stage, where the global branch is constructed to perceive deep-level features and the local branch is built to focus on the refined details. Furthermore, an encoder is built to generate some features, and a decoder is constructed to simulate decision behavior, followed by the information bottleneck viewpoint to optimize the objective. Extensive experiments are performed to evaluate our framework on two publicly available datasets, namely, 1) the Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) and 2) the Lung and Colon Histopathological Image Dataset (LC25000). For instance, our framework achieves 92.98% accuracy and presents additional visualizations on the LIDC. The experiment results show that our framework can obtain outstanding performance and is effective to facilitate explainability. It also demonstrates that this united framework is a serviceable tool and further has the scalability to be introduced into clinical research.

BackgroundEarly diagnosis of pulmonary nodules is essential for effective prevention and treatment of pulmonary cancer. However, the heterogeneous and complex characteristics of pulmonary nodules, such as shape, size, speculation, and texture, present significant challenges in clinical diagnosis, which computer-aided diagnosis (CAD) can help address. Moreover, the varied performance of deep learning methods in CAD and limited model interpretability often hinder clinicians’ understanding of CAD results. In this study, we propose a multitask Swin Transformer (MTST) for classifying benign and malignant pulmonary nodules, which outputs nodule features as classification criteria.MethodsWe introduce a MTST model for feature extraction, designed with a multitask layer that simultaneously outputs benign and malignant binary classification, multilevel classification, and a detailed analysis of pulmonary nodule features. In addition, we incorporate image augmentation using a U-Net generative adversarial network (GAN) model to enhance the training process.ResultsExperimental findings on the Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) dataset demonstrate that the proposed MTST outperforms conventional convolutional neural networks (CNN)-based networks across multiple tasks. Specifically, MTST achieved an accuracy of 93.24% in binary classification of benign and malignant nodules and demonstrated superior performance in nodule feature evaluation. For multilevel classification of pulmonary nodules, the Swin Transformer achieved an accuracy of 95.73%. On the training, validation, and test sets (9,600/2,400/1,600 nodules), the MTST model achieved an accuracy of 93.74%, sensitivity of 91.55%, and specificity of 96.09%. The results indicate that the MTST model aligns well with clinical diagnostic practices, offering improved performance and reliability.ConclusionsThe MTST model’s efficacy in binary classification, multiclass classification, and feature evaluation confirms its potential as a valuable tool for CAD systems in clinical settings.

Background Deep learning may help to improve computer-aided detection of volume (CADv) measurement of pulmonary nodules at chest CT. Purpose To determine the efficacy of a deep learning method for improving CADv for measuring the solid and ground-glass opacity (GGO) volumes of a nodule, doubling time (DT), and the change in volume at chest CT. Materials and Methods From January 2014 to December 2016, patients with pulmonary nodules at CT were retrospectively reviewed. CADv without and with a convolutional neural network (CNN) automatically determined total nodule volume change per day and DT. Area under the curves (AUCs) on a per-nodule basis and diagnostic accuracy on a per-patient basis were compared among all indexes from CADv with and without CNN for differentiating benign from malignant nodules. Results The CNN training set was 294 nodules in 217 patients, the validation set was 41 nodules in 32 validation patients, and the test set was 290 nodules in 188 patients. A total of 170 patients had 290 nodules (mean size ± standard deviation, 11 mm ± 5; range, 4-29 mm) diagnosed as 132 malignant nodules and 158 benign nodules. There were 132 solid nodules (46%), 106 part-solid nodules (36%), and 52 ground-glass nodules (18%). The test set results showed that the diagnostic performance of the CNN with CADv for total nodule volume change per day was larger than DT of CADv with CNN (AUC, 0.94 [95% confidence interval {CI}: 0.90, 0.96] vs 0.67 [95% CI: 0.60, 0.74]; P < .001) and CADv without CNN (total nodule volume change per day: AUC, 0.69 [95% CI: 0.62, 0.75]; P < .001; DT: AUC, 0.58 [95% CI: 0.51, 0.65]; P < .001). The accuracy of total nodule volume change per day of CADv with CNN was significantly higher than that of CADv without CNN (P < .001) and DT of both methods (P < .001). Conclusion Convolutional neural network is useful for improving accuracy of computer-aided detection of volume measurement and nodule differentiation capability at CT for patients with pulmonary nodules. © RSNA, 2020 Online supplemental material is available for this article.

Lung cancer is the leading cause of cancer related deaths in the world. The survival rate can be improved if the presence of lung nodules are detected early. This has also led to more focus being given to computer aided detection (CAD) and diagnosis of lung nodules. The arbitrariness of shape, size and texture of lung nodules is a challenge to be faced when developing these detection systems. In the proposed work we use convolutional neural networks to learn the features for nodule detection, replacing the traditional method of handcrafting features like geometric shape or texture. Our network uses the DetectNet architecture based on YOLO (You Only Look Once) to detect the nodules in CT scans of lung. In this architecture, object detection is treated as a regression problem with a single convolutional network simultaneously predicting multiple bounding boxes and class probabilities for those boxes. By performing training using chest CT scans from Lung Image Database Consortium (LIDC), NVIDIA DIGITS and Caffe deep learning framework, we show that nodule detection using this single neural network can result in reasonably low false positive rates with high sensitivity and precision.

Background: Lung cancer is the leading cause of cancer-related mortality globally. Early detection via low-dose computed tomography (LDCT) can reduce mortality, but its implementation is challenged by the absence of objective diagnostic criteria and the necessity for extensive manual interpretation. Public datasets like the Lung Image Database Consortium often lack pathology-confirmed diagnoses, which can lead to inaccuracies in ground truth labels. Variability in voxel sizes across these datasets also complicates feature extraction, undermining model reliability. Many existing methods for integrating nodule boundary annotations use deep learning models such as generative adversarial networks, which often lack interpretability. Methods: This study assesses the effect of voxel normalization on pulmonary nodule classification and introduces a Fast Fourier Transform-based contour fusion method as a more interpretable alternative. Utilizing pathology-confirmed LDCT data from 415 patients across two medical centers, both machine learning and deep learning models were developed using voxel-normalized images and attention mechanisms, including transformers. Results: The results demonstrated that voxel normalization significantly improved the overlap of features between datasets from two different centers by 64%, resulting in enhanced selection stability. In the ROI-based radiomics analysis, the top-performing machine-learning model achieved an accuracy of 92.6%, whereas the patch-based deep-learning models reached 98.5%. Notably, the FFT-based method provided a clinically interpretable integration of expert annotations, effectively addressing a major limitation of generative adversarial networks. Conclusions: Voxel normalization enhances reliability in pulmonary nodule classification while the FFT-based method offers a viable path toward interpretability in deep learning applications. Future research should explore its implications further in multi-center contexts.

Lung cancer is one of the most common malignant tumors leading to death. Early diagnosis of pulmonary nodules is of great significance for reducing the mortality of the lung cancer patients. Recently, deep learning, i.e., the most potential technology in computer vision, has been achieving brilliant performance in a variety of medical image tasks. In this paper, we explore a novel pulmonary nodule classification method based on the recent ResNet and 2.5D view of CT images. The 2.5D view images are constructed by the 2D images that are re-sampled from the original 3D CT images through three different planes, i.e., coronal plane, sagittal plane and axial plane. Then, the recent proposed ResNet-50 network is adopted for feature extraction and classification of the pulmonary nodules in an end-to-end manner. The given method is evaluated on the public LIDC database, and the comparative experimental results show its effectiveness for the classification of benign and malignant pulmonary nodules.

Pulmonary nodule classification from computerized tomography(CT) Scans is a vital task for the early screening of Lung cancers. The algorithm is aiming at distinguishing malignant pulmonary nodules, benign nodules and the ones with their subtypes. In this paper, we defined a detailed pulmonary nodule classification task considering 5 semantic labels. We are facing with a series of non-trival problems dealing with such a task. First, the available medical image data for training is quite limited. We enlarged the training dataset by cropping out three-dimension(3D) volume of each pulmonary nodule and generating 15 planes with different orientations from these volumes. Secondly, the global modeling ability of the existing convolutional neural network(CNN) based architectures can not meet the need of medical image analysis well. To learn discriminative abstract information, we down-sample feature maps between successive stages and adopt the BotNet-50 backbone which is a combination of ResNet backbone and self-attention modules. Such an architecture can extract local and non-local information in low-level and high-level layers, respectively. Last but not the least, the data distribution of training data and testing data don't share similar distribution in real-world multi-center medical image classification scenes. We assigned the samples with modified wights while calculating the loss value for optimization. The proposed method can eliminate the spurious correlation between features and labels. Experiments demonstrate the effectiveness of each component.

Classification of lung nodule malignancy in computed tomography imaging utilising generative adversarial networks and semi-supervised transfer learning