scholarly journals Multi-Level Cross Residual Network for Lung Nodule Classification

Sensors ◽  
2020 ◽  
Vol 20 (10) ◽  
pp. 2837 ◽  
Author(s):  
Juan Lyu ◽  
Xiaojun Bi ◽  
Sai Ho Ling
Keyword(s):  
Lung Cancer ◽  
Early Stage ◽  
Binary Classification ◽  
Lung Nodule ◽  
Disease Diagnosis ◽  
Lung Nodules ◽  
Lung Cancer Patients ◽  
Multi Scale ◽  
Multi Level ◽  
Nodule Classification

Computer-aided algorithm plays an important role in disease diagnosis through medical images. As one of the major cancers, lung cancer is commonly detected by computer tomography. To increase the survival rate of lung cancer patients, an early-stage diagnosis is necessary. In this paper, we propose a new structure, multi-level cross residual convolutional neural network (ML-xResNet), to classify the different types of lung nodule malignancies. ML-xResNet is constructed by three-level parallel ResNets with different convolution kernel sizes to extract multi-scale features of the inputs. Moreover, the residuals are connected not only with the current level but also with other levels in a crossover manner. To illustrate the performance of ML-xResNet, we apply the model to process ternary classification (benign, indeterminate, and malignant lung nodules) and binary classification (benign and malignant lung nodules) of lung nodules, respectively. Based on the experiment results, the proposed ML-xResNet achieves the best results of 85.88% accuracy for ternary classification and 92.19% accuracy for binary classification, without any additional handcrafted preprocessing algorithm.

scholarly journals Lung Cancer Classification Using Squeeze and Excitation Convolutional Neural Networks with Grad Cam++ Class Activation Function

2021 ◽  
Vol 38 (4) ◽  
pp. 1103-1112
Author(s):  
Eali Stephen Neal Joshua ◽  
Debnath Bhattacharyya ◽  
Midhun Chakkravarthy ◽  
Hye-Jin Kim
Keyword(s):  
Lung Cancer ◽  
Patient Survival ◽  
Lung Nodule ◽  
Activation Function ◽  
Cancer Classification ◽  
Systematic Evaluation ◽  
Lung Nodules ◽  
Nodule Detection ◽  
Lung Cancer Therapy ◽  
Nodule Classification

The leading cause of cancer-related death globally has been identified as lung cancer. Early lung nodule detection is critical for lung cancer therapy and patient survival. The Gard Cam++ Class Activation Function is used with a squeeze-and-excite network to provide a revolutionary method for differentiating malignant from benign lung nodules on CT scans. The new SENET (Squeeze-and-Excitation Networks) Grad Cam++ module, which combines the features calibration and discrimination benefits of SENET, has been shown to have a substantial potential for improving feature discriminability in lung cancer classification. According to the publicly available LUng Nodule Analysis 2016 (LUNA16) database, when assessed on 1230 nodules, the technique achieved an AUC of 0.9664 and an accuracy of 97.08% (600 malignant and 630 benign). The favorable results demonstrate the robustness of our technique to nodule classification, which we anticipate will be valuable in the future. The technology's objective is to aid radiologists in evaluating diagnostic data and differentiating benign from malignant lung nodules on CT images. To our knowledge, no systematic evaluation of SENET usefulness in classifying lung nodules has been done.

scholarly journals Segmentation of Lung Nodules Using Improved 3D-UNet Neural Network

Symmetry ◽  
2020 ◽  
Vol 12 (11) ◽  
pp. 1787
Author(s):  
Zhitao Xiao ◽  
Bowen Liu ◽  
Lei Geng ◽  
Fang Zhang ◽  
Yanbei Liu
Keyword(s):  
Neural Network ◽  
Lung Cancer ◽  
Image Segmentation ◽  
Deep Level ◽  
Lung Nodule ◽  
Recall Rate ◽  
Lung Nodules ◽  
Multi Scale ◽  
Diagnosis Of Lung Cancer ◽  
Computed Tomography Image

Lung cancer has one of the highest morbidity and mortality rates in the world. Lung nodules are an early indicator of lung cancer. Therefore, accurate detection and image segmentation of lung nodules is of great significance to the early diagnosis of lung cancer. This paper proposes a CT (Computed Tomography) image lung nodule segmentation method based on 3D-UNet and Res2Net, and establishes a new convolutional neural network called 3D-Res2UNet. 3D-Res2Net has a symmetrical hierarchical connection network with strong multi-scale feature extraction capabilities. It enables the network to express multi-scale features with a finer granularity, while increasing the receptive field of each layer of the network. This structure solves the deep level problem. The network is not prone to gradient disappearance and gradient explosion problems, which improves the accuracy of detection and segmentation. The U-shaped network ensures the size of the feature map while effectively repairing the lost features. The method in this paper was tested on the LUNA16 public dataset, where the dice coefficient index reached 95.30% and the recall rate reached 99.1%, indicating that this method has good performance in lung nodule image segmentation.

scholarly journals Transthoracic fine-needle aspiration diagnosis of solid, subsolid, and partially calcified lung nodules: A retrospective study from a single academic center

CytoJournal ◽  
2019 ◽  
Vol 16 ◽  
pp. 16 ◽  
Author(s):  
Yangying Zhou ◽  
Gary Gong ◽  
Haiyan Wang ◽  
Zahra Alikhassy Habibabady ◽  
Peggy Lang ◽  
...  
Keyword(s):  
Lung Cancer ◽  
Fine Needle Aspiration ◽  
Large Scale ◽  
Early Stage ◽  
Lung Nodule ◽  
Needle Aspiration ◽  
Lung Nodules ◽  
Computer Search ◽  
Radiographic Findings ◽  
Fine Needle

Background: The large-scale National Lung Cancer Screening Trial demonstrated an increased detection of early-stage lung cancers using low-dose computed tomography scan in the screening population. It also demonstrated a 20% reduction of lung cancer-related deaths in these patients. Aims: Although both solid and subsolid lung nodules are evaluated in studies, subsolid and partially calcified lung nodules are often overlooked. Materials and Methods: We reviewed transthoracic fine-needle aspiration (FNA) cases from lung nodule patients in our clinics and correlated cytological diagnoses with radiologic characteristics of lesions. A computer search of the pathology archive was performed over a period of 12 months for transthoracic FNAs, including both CT- and ultrasound-guided biopsies. Results: A total of 111 lung nodule cases were identified. Lesions were divided into three categories: solid, subsolid, and partially calcified nodules according to radiographic findings. Of 111 cases, the average sizes of the solid (84 cases), subsolid (22 cases), and calcified (5 cases) lesions were 1.952 ± 2.225, 1.333 ± 1.827, and 1.152 ± 1.984 cm, respectively. The cytological diagnoses of three groups were compared. A diagnosis of malignancy was made in 64.28% (54 cases) in solid, 22.72% (5 cases) in subsolid, and 20% (1 case) in partially calcified nodules. Among benign lesions, eight granulomatous inflammations were identified, including one case of solid, five cases of subsolid, and two cases of calcified nodules. Conclusions: Our study indicates that solid nodules have the highest risk of malignancy. Furthermore, the cytological evaluation of subsolid and partially calcified nodules is crucial for the accurate diagnosis and appropriate clinical management of lung nodule patients.

scholarly journals Development and outcomes of a comprehensive multidisciplinary incidental lung nodule and lung cancer screening program

2019 ◽  
Author(s):  
Gregory LeMense ◽  
Ernest A. Waller ◽  
Cheryl Campbell ◽  
Tyler Bowen
Keyword(s):  
Lung Cancer ◽  
Cancer Screening ◽  
Cancer Diagnosis ◽  
Screening Program ◽  
Early Stage ◽  
Lung Nodule ◽  
Lung Cancer Screening ◽  
Lung Nodules ◽  
Hospital System ◽  
Number Of Patients

Abstract BackgroundAppropriate management of lung nodules detected incidentally or through lung cancer screening can increase the rate of early-stage diagnoses and potentially improve treatment outcomes. However, the implementation and management of comprehensive lung nodule programs is challenging. MethodsA single-center, retrospective study was conducted to describe the development and outcomes of a lung nodule program at a community practice in Tennessee.ResultsThe number of patients with lung nodules referred to the program increased over 2 years, with 665 patients in Year 1 and 745 patients in Year 2. Most nodules were incidental (60% Year 1, 65% Year 2). In Year 1, 17% of nodules were symptomatic and 12% were identified through screening. Of the 665 nodules in Year 1, 182 underwent a diagnostic intervention and 121 (18%) received a cancer diagnosis. Most diagnostic interventions were image-guided bronchoscopy (88%) or percutaneous biopsy (9%). The proportion of Stage I-II cancer diagnoses increased from 23% prior to program implementation to 36% in Year 1 and 38% in Year 2. Among screening cases, follow-up scans were conducted within 18 months in 71%. Only 2% of patients under watchful waiting required a diagnostic intervention, of which 1% received a cancer diagnosis.ConclusionsThe current study reports outcomes over the first two years of a lung cancer screening and incidental nodule program. The program was successful and manageable, given the appropriate level of data management and oversight. Comprehensive lung nodule programs have the potential to benefit the patient, physician, and hospital system.

scholarly journals Feasibility Study of MRI and Multimodality CT/MRI Radiomics for Lung Nodule Classification

2021 ◽  
Author(s):  
Anthony E. A. Jatobá ◽  
Marcelo C. Oliveira ◽  
Marcel Koenigkam-Santos ◽  
Paulo de Azevedo-Marques
Keyword(s):  
Lung Cancer ◽  
Lung Nodule ◽  
Lung Cancer Screening ◽  
Machine Learning Algorithms ◽  
Lung Nodules ◽  
Mri Features ◽  
The Individual ◽  
Multimodality Approach ◽  
Nodule Classification ◽  
Ct And Mri

Lung cancer is the most common and lethal form of cancer, and its early diagnosis is key to the patient's survival. CT is the reference imaging scan for lung cancer screening; however, it presents the drawback of exposing the patient to ionizing radiation. Recent studies have shown the relevance of MRI in lung nodules diagnosis. In this work, we aimed to evaluate whether radiomics features from MRI are well-suited for lung nodules characterization and if the combination of CT and MRI features can yield better results than the features from the individual modalities. For such, we segmented paired CT and MRI nodules from 33 lung nodules patients, extracted 89 radiomics features from each modality, and combined it into a multimodality feature set. Those features were then used for classifying the nodules into benign and malignant by a set of machine learning algorithms, assessing the AUC across 30 trials. Our results show that MRI radiomics features are suitable for characterizing lung lesions, yielding AUC values up to 17% higher than their CT counterparts, and shedding light on MRI as a viable image modality for decision support systems. Conversely, our multimodality approach did not improve performance compared to the single-modality models, suggesting that the direct combination of multimodality features might not be an adequate strategy for dealing with multimodality medical images.

Multi-level 3D Densenets for False-positive Reduction in Lung Nodule Detection Based on Chest Computed Tomography

2020 ◽  
Vol 16 (8) ◽  
pp. 1004-1021
Author(s):  
Xiaoqi Lu ◽  
Yu Gu ◽  
Lidong Yang ◽  
Baohua Zhang ◽  
Ying Zhao ◽  
...  
Keyword(s):  
False Positive ◽  
Contextual Information ◽  
Lung Nodule ◽  
Positive Sample ◽  
Lung Nodules ◽  
Lung Nodule Detection ◽  
Nodule Detection ◽  
Level Information ◽  
Multi Level ◽  
Cade System

Objective: False-positive nodule reduction is a crucial part of a computer-aided detection (CADe) system, which assists radiologists in accurate lung nodule detection. In this research, a novel scheme using multi-level 3D DenseNet framework is proposed to implement false-positive nodule reduction task. Methods: Multi-level 3D DenseNet models were extended to differentiate lung nodules from falsepositive nodules. First, different models were fed with 3D cubes with different sizes for encoding multi-level contextual information to meet the challenges of the large variations of lung nodules. In addition, image rotation and flipping were utilized to upsample positive samples which consisted of a positive sample set. Furthermore, the 3D DenseNets were designed to keep low-level information of nodules, as densely connected structures in DenseNet can reuse features of lung nodules and then boost feature propagation. Finally, the optimal weighted linear combination of all model scores obtained the best classification result in this research. Results: The proposed method was evaluated with LUNA16 dataset which contained 888 thin-slice CT scans. The performance was validated via 10-fold cross-validation. Both the Free-response Receiver Operating Characteristic (FROC) curve and the Competition Performance Metric (CPM) score show that the proposed scheme can achieve a satisfactory detection performance in the falsepositive reduction track of the LUNA16 challenge. Conclusion: The result shows that the proposed scheme can be significant for false-positive nodule reduction task.

Lung nodule classification using combination of CNN, second and higher order texture features

2021 ◽  
pp. 1-9
Author(s):  
Amrita Naik ◽  
Damodar Reddy Edla
Keyword(s):  
Malignant Tumors ◽  
Early Stage ◽  
Lung Nodule ◽  
Texture Features ◽  
Svm Classifier ◽  
Softmax Classifier ◽  
Spatial Features ◽  
High Level ◽  
Learning Architectures ◽  
Nodule Classification

Lung cancer is the most common cancer throughout the world and identification of malignant tumors at an early stage is needed for diagnosis and treatment of patient thus avoiding the progression to a later stage. In recent times, deep learning architectures such as CNN have shown promising results in effectively identifying malignant tumors in CT scans. In this paper, we combine the CNN features with texture features such as Haralick and Gray level run length matrix features to gather benefits of high level and spatial features extracted from the lung nodules to improve the accuracy of classification. These features are further classified using SVM classifier instead of softmax classifier in order to reduce the overfitting problem. Our model was validated on LUNA dataset and achieved an accuracy of 93.53%, sensitivity of 86.62%, the specificity of 96.55%, and positive predictive value of 94.02%.

scholarly journals An Efficient DA-Net Architecture for Lung Nodule Segmentation

Mathematics ◽  
2021 ◽  
Vol 9 (13) ◽  
pp. 1457
Author(s):  
Muazzam Maqsood ◽  
Sadaf Yasmin ◽  
Irfan Mehmood ◽  
Maryam Bukhari ◽  
Mucheol Kim
Keyword(s):  
Computed Tomography ◽  
Lung Cancer ◽  
Ct Scan ◽  
Cancer Diagnosis ◽  
Lung Nodule ◽  
Search Space ◽  
Lung Nodules ◽  
Suggested Model ◽  
Lung Cancer Diagnosis ◽  
Segmentation Framework

A typical growth of cells inside tissue is normally known as a nodular entity. Lung nodule segmentation from computed tomography (CT) images becomes crucial for early lung cancer diagnosis. An issue that pertains to the segmentation of lung nodules is homogenous modular variants. The resemblance among nodules as well as among neighboring regions is very challenging to deal with. Here, we propose an end-to-end U-Net-based segmentation framework named DA-Net for efficient lung nodule segmentation. This method extracts rich features by integrating compactly and densely linked rich convolutional blocks merged with Atrous convolutions blocks to broaden the view of filters without dropping loss and coverage data. We first extract the lung’s ROI images from the whole CT scan slices using standard image processing operations and k-means clustering. This reduces the search space of the model to only lungs where the nodules are present instead of the whole CT scan slice. The evaluation of the suggested model was performed through utilizing the LIDC-IDRI dataset. According to the results, we found that DA-Net showed good performance, achieving an 81% Dice score value and 71.6% IOU score.

Sign in / Sign up

Export Citation Format

Share Document