Automatic Nodule Detection Research Articles

Establishment and preliminary application of personalized three-dimensional reconstruction of thyroid gland with automatic detection of thyroid nodules based on ultrasound videos.

A well display of the spatial location of thyroid nodules in the thyroid is important for surgical path planning and surgeon-patient communication. The aim of this study was to establish a three-dimensional (3D) reconstruction method of the thyroid gland, thyroid nodule, and carotid artery with automatic detection based on two-dimensional (2D) ultrasound videos, and to evaluate its clinical value. Ultrasound videos, including the thyroid gland with nodule, isthmus of thyroid gland, and ipsilateral carotid artery, were recorded. BC-UNet, MTN-Net, and RDPA-U-Net network models were innovatively employed for segmentation of the thyroid glands, the thyroid nodules, and the carotid artery respectively. Marching Cubes algorithm was used for reconstruction, while Laplacian smoothing algorithm was employed to smooth the 3D model surface. Using this model, 20 patients and 15 surgeons completed surveys on the effectiveness of this model for the pre-surgery demonstration of nodule location as well as surgeon-patient communication. The thyroid gland with nodule, isthmus of gland, and carotid artery were reconstructed and displayed. With the 3D model, the understanding of the spatial location of thyroid nodules improved in all three surgeon groups, eliminating the influence of professional levels. In the patient survey, the patients' understanding of the thyroid nodule location and procedure for surgery were significantly improved. In addition, with the 3D model, the time for doctors to explain to patients was significantly reduced (16.75 vs. 8.85min, p=0.001). To our knowledge, this is the first report of constructing a 3D thyroid model using a deep learning technique for personalized thyroid segmentation based on 2D ultrasound videos. The preliminary clinical application showed that it was conducive to the comprehension of the location of thyroid nodules for surgeons and patients, with significant improvement on the efficiency of surgeon-patient communication.

Automatic detection of thyroid nodules with a real-time artificial intelligence system in a real clinical scenario and the associated influencing factors.

At present, most articles mainly focused on the diagnosis of thyroid nodules by using artificial intelligence (AI), and there was little research on the detection performance of AI in thyroid nodules. To explore the value of a real-time AI based on computer-aided diagnosis system in the detection of thyroid nodules and to analyze the factors influencing the detection accuracy. From June 1, 2022 to December 31, 2023, 224 consecutive patients with 587 thyroid nodules were prospective collected. Based on the detection results determined by two experienced radiologists (both with more than 15 years experience in thyroid diagnosis), the detection ability of thyroid nodules of radiologists with different experience levels (junior radiologist with 1 year experience and senior radiologist with 5 years experience in thyroid diagnosis) and real-time AI were compared. According to the logistic regression analysis, the factors influencing the real-time AI detection of thyroid nodules were analyzed. The detection rate of thyroid nodules by real-time AI was significantly higher than that of junior radiologist (P = 0.013), but lower than that of senior radiologist (P = 0.001). Multivariate logistic regression analysis showed that nodules size, superior pole, outside (near carotid artery), close to vessel, echogenicity (isoechoic, hyperechoic, mixed-echoic), morphology (not very regular, irregular), margin (unclear), ACR TI-RADS category 4 and 5 were significant independent influencing factors (all P < 0.05). With the combination of real-time AI and radiologists, junior and senior radiologist increased the detection rate to 97.4% (P < 0.001) and 99.1% (P = 0.015) respectively. The real-time AI has good performance in thyroid nodule detection and can be a good auxiliary tool in the clinical work of radiologists.

Fully automatic deep learning-based lung parenchyma segmentation and boundary correction in thoracic CT scans.

The proposed work aims to develop an algorithm to precisely segment the lung parenchyma in thoracic CT scans. To achieve this goal, the proposed technique utilized a combination of deep learning and traditional image processing algorithms. The initial step utilized a trained convolutional neural network (CNN) to generate preliminary lung masks, followed by the proposed post-processing algorithm for lung boundary correction. First, the proposed method trained an improved 2D U-Net CNN model with Inception-ResNet-v2 as its backbone. The model was trained on 32 CT scans from two different sources: one from the VESSEL12 grand challenge and the other from AIIMS Delhi. Further, the model's performance was evaluated on a test dataset of 16 CT scans with juxta-pleural nodules obtained from AIIMS Delhi and the LUNA16 challenge. The model's performance was assessed using evaluation metrics such as average volumetric dice coefficient (DSCavg), average IoU score (IoUavg), and average F1 score (F1avg). Finally, the proposed post-processing algorithm was implemented to eliminate false positives from the model's prediction and to include juxta-pleural nodules in the final lung masks. The trained model reported a DSCavg of 0.9791 ± 0.008, IoUavg of 0.9624 ± 0.007, and F1avg of 0.9792 ± 0.004 on the test dataset. Applying the post-processing algorithm to the predicted lung masks obtained a DSCavg of 0.9713 ± 0.007, IoUavg of 0.9486 ± 0.007, and F1avg of 0.9701 ± 0.008. The post-processing algorithm successfully included juxta-pleural nodules in the final lung mask. Using a CNN model, the proposed method for lung parenchyma segmentation produced precise segmentation results. Furthermore, the post-processing algorithm addressed false positives and negatives in the model's predictions. Overall, the proposed approach demonstrated promising results for lung parenchyma segmentation. The method has the potential to be valuable in the advancement of computer-aided diagnosis (CAD) systems for automatic nodule detection.

Performance of alternative manual and automated deep learning segmentation techniques for the prediction of benign and malignant lung nodules.

We aim to evaluate the performance of radiomic biopsy (RB), best-fit bounding box (BB), and a deep-learning-based segmentation method called no-new-U-Net (nnU-Net), compared to the standard full manual (FM) segmentation method for predicting benign and malignant lung nodules using a computed tomography (CT) radiomic machine learning model. A total of 188 CT scans of lung nodules from 2 institutions were used for our study. One radiologist identified and delineated all 188 lung nodules, whereas a second radiologist segmented a subset () of these nodules. Both radiologists employed FM and RB segmentation methods. BB segmentations were generated computationally from the FM segmentations. The nnU-Net, a deep-learning-based segmentation method, performed automatic nodule detection and segmentation. The time radiologists took to perform segmentations was recorded. Radiomic features were extracted from each segmentation method, and models to predict benign and malignant lung nodules were developed. The Kruskal-Wallis and DeLong tests were used to compare segmentation times and areas under the curve (AUC), respectively. For the delineation of the FM, RB, and BB segmentations, the two radiologists required a median time (IQR) of 113 (54 to 251.5), 21 (9.25 to 38), and 16 (12 to 64.25) s, respectively (). In dataset 1, the mean AUC (95% CI) of the FM, RB, BB, and nnU-Net model were 0.964 (0.96 to 0.968), 0.985 (0.983 to 0.987), 0.961 (0.956 to 0.965), and 0.878 (0.869 to 0.888). In dataset 2, the mean AUC (95% CI) of the FM, RB, BB, and nnU-Net model were 0.717 (0.705 to 0.729), 0.919 (0.913 to 0.924), 0.699 (0.687 to 0.711), and 0.644 (0.632 to 0.657). Radiomic biopsy-based models outperformed FM and BB models in prediction of benign and malignant lung nodules in two independent datasets while deep-learning segmentation-based models performed similarly to FM and BB. RB could be a more efficient segmentation method, but further validation is needed.

MANet: Multi-branch attention auxiliary learning for lung nodule detection and segmentation

Background and objective: Pulmonary nodule detection and segmentation are currently two primary tasks in analyzing chest computed tomography (Chest CT) in order to detect signs of lung cancer, thereby providing early treatment measures to reduce mortality. Even though there are many proposed methods to reduce false positives for obtaining effective detection results, distinguishing between the pulmonary nodule and background region remains challenging because their biological characteristics are similar and varied in size. The purpose of our work is to propose a method for automatic nodule detection and segmentation in Chest CT by enhancing the feature information of pulmonary nodules.Methods: We propose a new UNet-based backbone with multi-branch attention auxiliary learning mechanism, which contains three novel modules, namely, Projection module, Fast Cascading Context module, and Boundary Enhancement module, to further enhance the nodule feature representation. Based on that, we build MANet, a lung nodule localization network that simultaneously detects and segments precise nodule positions. Furthermore, our MANet contains a Proposal Refinement step which refines initially generated proposals to effectively reduce false positives and thereby produce the segmentation quality.Results: Comprehensive experiments on the combination of two benchmarks LUNA16 and LIDC-IDRI show that our proposed model outperforms state-of-the-art methods in the tasks of nodule detection and segmentation tasks in terms of FROC, IoU, and DSC metrics. Our method reports an average FROC score of 88.11% in lung nodule detection. For the lung nodule segmentation, the results reach an average IoU score of 71.29% and a DSC score of 82.74%. The ablation study also shows the effectiveness of the new modules which can be integrated into other UNet-based models.Conclusions: The experiments demonstrated our method with multi-branch attention auxiliary learning ability are a promising approach for detecting and segmenting the pulmonary nodule instances compared to the original UNet design.

Computer-Aided Detection of Subsolid Nodules on Chest Computed Tomography: Assessment of Visualization on Vessel-Suppressed Images.

This study aimed to clarify the performance of automatic detection of subsolid nodules by commercially available software on computed tomography (CT) images of various slice thicknesses and compare it with visualization on the accompanying vessel-suppression CT (VS-CT) images. A total of 95 subsolid nodules from 84 CT examinations of 84 patients were included. The reconstructed CT image series of each case with 3-, 2-, and 1-mm slice thicknesses were loaded into a commercially available software application (ClearRead CT) for automatic detection of subsolid nodules and generation of VS-CT images. Automatic nodule detection sensitivity was assessed for 95 nodules on each series of images acquired at 3 slice thicknesses. Four radiologists subjectively evaluated visual assessment of the nodules on VS-CT. ClearRead CT automatically detected 69.5% (66/95 nodules), 68.4% (65/95 nodules), and 70.5% (67/95 nodules) of all subsolid nodules in 3-, 2-, and 1-mm slices, respectively. The detection rate was higher for part-solid nodules than for pure ground-glass nodules at all slice thicknesses. In the visualization assessment on VS-CT, 3 nodules at each slice thickness (3.2%) were judged as invisible, while 26 of 29 (89.7%), 27 of 30 (90.0%), and 25 of 28 (89.3%) nodules, which were missed by computer-aided detection, were judged as visible in 3-, 2-, and 1-mm slices, respectively. The automatic detection rate of subsolid nodules by ClearRead CT was approximately 70% at all slice thicknesses. More than 95% of subsolid nodules were visualized on VS-CT, including nodules undetected by the automated software. Computed tomography acquisition at slices thinner than 3 mm did not confer any benefits.

Improved UNet Deep Learning Model for Automatic Detection of Lung Cancer Nodules.

Uncontrolled cell growth in the two spongy lung organs in the chest is the most prevalent kind of cancer. When cells from the lungs spread to other tissues and organs, this is referred to as metastasis. This work uses image processing, deep learning, and metaheuristics to identify cancer in its early stages. At this point, a new convolutional neural network is constructed. The predator technique has the potential to increase network architecture and accuracy. Deep learning identified lung cancer spinal metastases in as energy consumption increased CT readings for lung cancer bone metastases decreased. Qualified physicians, on the other hand, discovered 71.14 and 74.60 percent of targets with energies of 140 and 60 keV, respectively, whereas the proposed model gives 76.51 and 81.58 percent, respectively. Expert physicians' detection rate was 74.60 percent lower than deep learning's detection rate of 81.58 percent. The proposed method has the highest accuracy, sensitivity, and specificity (93.4, 98.4, and 97.1 percent, respectively), as well as the lowest error rate (1.6 percent). Finally, in lung segmentation, the proposed model outperforms the CNN model. High-intensity energy-spectral CT images are more difficult to segment than low-intensity energy-spectral CT images.

Lung Cancer Detection Using Modified AlexNet Architecture and Support Vector Machine

Lung cancer is the most dangerous and death-causing disease indicated by the presence of pulmonary nodules in the lung. It is mostly caused by the instinctive growth of cells in the lung. Lung nodule detection has a significant role in detecting and screening lung cancer in Computed tomography (CT) scan images. Early detection plays an important role in the survival rate and treatment of lung cancer patients. Moreover, pulmonary nodule classification techniques based on the convolutional neural network can be used for the accurate and efficient detection of lung cancer. This work proposed an automatic nodule detection method in CT images based on modified AlexNet architecture and Support vector machine (SVM) algorithm namely LungNet-SVM. The proposed model consists of seven convolutional layers, three pooling layers, and two fully connected layers used to extract features. Support vector machine classifier is applied for the binary classification of nodules into benign and malignant. The experimental analysis is performed by using the publicly available benchmark dataset Lung nodule analysis 2016 (LUNA16). The proposed model has achieved 97.64% of accuracy, 96.37% of sensitivity, and 99.08% of specificity. A comparative analysis has been carried out between the proposed LungNet-SVM model and existing state-of-the-art approaches for the classification of lung cancer. The experimental results indicate that the proposed LungNet-SVM model achieved remarkable performance on a LUNA16 dataset in terms of accuracy.

S25 Deep Learning for the Automatic Identification of Neoplastic Biliary Nodules in Patients With Indeterminate Biliary Stenosis During Digital Cholangioscopy

Introduction: Digital single-operator cholangioscopy (DSOC) provides direct visual exploration of the biliary tract. This tool has become essential for distinguishing benign from malignant biliary strictures. The visual aspect of these lesions is highly sensitive for the diagnosis of malignancy. Macroscopic characteristic such as tumor vessels, papillary projections, masses and nodules are common findings in patients with malignant biliary strictures. Recent endoscopic literature has demonstrated large interest in artificial intelligence (AI) systems for detection of endoscopic lesions. However, the application of these systems for automatic characterization of biliary lesions has not been explored. This study aimed to develop a convolutional neural network for automatic detection of biliary nodules in D-SOC images. Methods: A convolutional neural network (CNN) was designed for automatic identification of biliary nodules using DSOC images. A total of 16150 frames were extracted from a pool of 85 patients undergoing DSOC (Spyglass DS II system, Boston Scientific, Marlborough, MA, USA). Each frame was classified by two endoscopists with experience in D-SOC regarding the presence or absence of nodules. Two image datasets were built for training and validation of the CNN (80% and 20% of the full image dataset, respectively). The performance of the model was measured by calculating the area under the receiving operating characteristic curve (AUC), sensitivity, specificity, positive and negative predictive values (PPV and NPV, respectively). Results: The architecture of the network was optimized for the detection of malignant biliary nodules. The CNN reached a sensitivity of 94.0%, a specificity of 99.9%, a PPV of 98.8% and a NPV of 99.7%. The overall accuracy of the deep learning system was 99.6%. The AUC was 1.00. Conclusion: The development of deep learning algorithms for application to DSOC may further potentiate the diagnostic capabilities of this modality. The application of these AI systems on real-time may help to guide biopsies and, thus, mitigate the current limitations of DSOC-guided biopsies, to achieve a more accurate diagnosis and timelier treatments. Our deep learning algorithm has demonstrated high performance levels for the detection of malignant nodules of the biliary tract, with high sensitivity, specificity and overall accuracy (Figure 1).Figure 1.: 1A - Output obtained from the application of the CNN. A blue bar represents a correct prediction. B – benign/normal findings; NN – nodules. 1B - ROC analyses of the network’s performance. AUC: area under the curve; CNN: convolutional neural network; B: benign/normal findings; NN: nodules.

An efficient IISH-2D DCNN-based lung nodule classification using CT scan images

Lung cancer has been identified as the world’s leading cause of death. Precise detection and testing of lung nodules at an early stage are essential for the diagnosis of lung cancer, therefore, it is one of the best ways to prevent deaths of lung cancer. An automated nodule detection system provides a second opinion to radiologists during early diagnosis. Much existing research uses Deep Convolutional Networks (DCNNs) for lung nodules classification. However, DCNNs normally need careful tuning of hyperparameters to reveal their excellent performance. Although, with the increasing size of state-of-the-art convolutional neural networks, the evaluation cost of the traditional optimization algorithms has become deplorable in most cases. Also, Lung CT scans data have a data imbalance problem inherently. So, to solve those problems, this paper proposes IISH-2D DCNN for lung nodule classification. The proposed methodology consists of pre-processing and classification phases. In the pre-processing phase, the 3D-CT scan slice is converted into a 2D-slice, and then the nodule boundary is extracted by calculation of ROI. After that, the extracted boundary is given as input to the IISH-2D DCNN that classifies the lung nodules. The performance of the proposed methodology is compared to the existing works based on accuracy, sensitivity, and specificity metrics. Thus, the proposed model outperforms existing lung nodule classification methodologies with higher accuracy, sensitivity, and specificity that are 99.8%, 97%, and 99%, respectively. Also, the proposed methodology has fewer errors than the state-of-the-art methods. Hence, the suggested method attains better performance in lung nodule classification and proves to be more effective.

Diagnosis of Pulmonary Nodules on CT Images Using YOLOv4

In this paper, the Scale-Invariant Feature Transform (SIFT) and Fast Library for Approximate Nearest Neighbors (FLANN) based algorithm is used to detect the abnormalities in the National Lung Screening Trial (NLST) CT scans as the exact clinical nodule locations are not provided in the dataset. These identified nodules on NLST CT Scans are then annotated using LabelImg tool. This process consumes time and so furthermore, the automatic nodule detection, You Only Look Once version 4 (YOLOv4) object detection model is implemented. The YOLOv4 object detection model is provided with total of 4187 labelled images in form of training (70%), validation (20%), and test (10%) datasets. Our YOLOv4 model achieves precision of 95%, sensitivity of 81% and mean Average Precision (mAP) of 89.1%.

Impact of dose reduction and iterative reconstruction algorithm on the detectability of pulmonary nodules by artificial intelligence

PurposeThe purpose of this study was to assess whether the performances of an automated software for lung nodule detection with computed tomography (CT) are affected by radiation dose and the use of iterative reconstruction algorithm. Materials and methodsA chest phantom (Multipurpose Chest Phantom N1; Kyoto Kagaku Co. Ltd, Kyoto, Japan) with 15 pulmonary nodules was scanned with a total of five CT protocol settings with up to 20-fold dose reduction. All CT examinations were reconstructed with iterative reconstruction algorithms ADMIRE 3 and ADMIRE 5 and were then analyzed for the presence of pulmonary nodules with a fully automated computer aided detection software system (InferReadTM CT Lung, Infervision), which is based on deep neural networks. ResultsThe sensitivity of fully automated pulmonary nodule detection for ground-glass nodules at standard dose CT was greater (70.0%; 14/20; 95% CI: 51.6-88.4%) than at 10-fold and 20-fold dose reduction (30.0%; 6/20; 95% CI: 0.0%-62.5%). There were less false positive findings when ADMIRE 5 reconstruction was used (4.0 ± 2.8 [SD]; range: 2–6) instead of ADMIRE 3 reconstruction (25.0 ± 15.6 [SD]; range: 14–36). There was no difference in the sensitivity of detection of solid and subsolid nodules between standard dose (100%; 95% CI: 100–100%) and 10- and 20-fold reduced dose CT (92.5%; 95% CI: 83.8–100.0%). Image noise was significantly greater with ADMIRE 3 (81 ± 2 [SD] [range: 79–84]; 104 ± 3 [SD] [range: 101–107]; 114 ± 5 [SD] [range: 110–119]; 193 ± 10 [SD] [range: 183-203]; 220 ± 16 [SD] [range: 210–238]) compared to ADMIRE 5 (44 ± 2 [SD] [range: 42–46]; 60 ± 2 [SD] [range: 57–61]; 66 ± 1 [SD] [range: 65–67]; 103 ± 4 [SD] [range: 98–106]; 110 ± 1 [SD] [range: 109–111]), respectively in each of the five CT protocols. ConclusionThis phantom study suggests that dose reduction and iterative reconstruction settings have an impact on detectability of pulmonary nodules by artificial intelligence software and we therefore encourage adaption of dose levels and reconstruction methods prior to widespread implementation of fully automatic nodule detection software for lung cancer screening purposes.

A deep learning-based method for detecting and classifying the ultrasound images of suspicious thyroid nodules.

The incidence of thyroid cancer has significantly increased in the last few decades. However, diagnosis of the thyroid nodules is labor and time intensive for radiologists and strongly depends on the personal experience of the radiologists. In this pursuit, the present study envisaged to develop a deep learning-based computer-aided diagnosis (CAD) method that enabled the automatic detection and classification of suspicious thyroid nodules in order to reduce the unnecessary fine-needle aspiration biopsy. The CAD method consisted of two main parts: detecting the location of thyroid nodules using a multiscale detection network and classifying the detected thyroid nodules by an attention-based classification network. The performance of the proposed method was evaluated and compared with that of other state-of-the-art deep learning methods and experienced radiologists. The proposed detection method outperformed three other detection architectures (average precision, 82.1%vs. 78.3%, 77.2%, and 74.8%). Moreover, the classification method showed a superior performance compared with four other state-of-the-art classification networks (accuracy, 94.8%vs. 91.2%, 85.0%, 80.8%, and 72.1%) and that by experienced radiologists (mean value of area under the curve, 0.941vs. 0.833). Our study verified the high efficiency of the proposed detection method. The findings can help improve the diagnostic performance of radiologists. However, the developed CAD system requires more training and evaluation in a large-population study.

Automatic detection of thyroid nodules with ultrasound images: Basing on semi-supervised learning

Ultrasonography is one of the most common method for the diagnosis of thyroid nodules clinically. But the lack of experienced radiologists has become the biggest problem which reduce the accuracy and effectiveness of thyroid nodules’ detection. What’s more, although artificial intelligence (AI) has been widely used to solve the problem, the performance is limited by the size of the database. In this paper, an improved U-Net is designed for the segmentation of thyroid nodules, following a fully convolution network for the classification of thyroid nodules. During the training, pseudo-label is used for semi-supervised learning. The whole work is trained in 2032 ultrasound images (1000 nodules) and tested in 400 ultrasound images (200 nodules). The average dice coefficient is 89.7% in the test set for segmentation, while the precision and recall is 93.2% and 96% separately for classification. These results are higher than the method which only use the images with true labels, proving the effectiveness of pseudo-label.

A cascade and heterogeneous neural network for CT pulmonary nodule detection and its evaluation on both phantom and patient data.

Screening of pulmonary nodules in computed tomography (CT) is crucial for early diagnosis and treatment of lung cancer. Although computer-aided diagnosis (CAD) systems have been designed to assist radiologists to detect nodules, fully automated detection is still challenging due to variations in nodule size, shape, and density. In this paper, we first propose a fully automated nodule detection method using a cascade and heterogeneous neural network trained on chest CT images of 12155 patients, then evaluate the performance by using phantom (828 CT images) and clinical datasets (2640 CT images) scanned with different imaging parameters. The nodule detection network employs two feature pyramid networks (FPNs) and a classification network (BasicNet). The first FPN is trained to achieve high sensitivity for nodule detection, and the second FPN refines the candidates for false positive reduction (FPR). Then, a BasicNet is combined with the second FPR to classify the candidates into either nodules or non-nodules for the final refinement. This study investigates the performance of nodule detection of solid and ground-glass nodules in phantom and patient data scanned with different imaging parameters. The results show that the detection of the solid nodules is robust to imaging parameters, and for GGO detection, reconstruction methods "iDose4-YA" and "STD-YA" achieve better performance. For thin-slice images, higher performance is achieved across different nodule sizes with reconstruction method "iDose4-STD". For 5 mm slice thickness, the best choice is the reconstruction method "iDose4-YA" for larger nodules (>5 mm). Overall, the reconstruction method "iDose4-YA" is suggested to achieve the best balanced results for both solid and GGO nodules.

Lung Nodule Classification on Computed Tomography Images Using Fractalnet

Lung Cancer is the most common cancer all over the world and is mostly diagnosed at later stages, thus increasing the risk of death. For early detection of malignant nodules there is a need for automated nodule detection system using imaging modalities like CT scan of lungs. Several machine learning algorithm specially deep learning algorithm have been used to accomplish this task, but without guaranteed accuracy. In the proposed work, we intend to improve the accuracy of pulmonary nodule classification system using fractalnet architecture. Fractalnet is also compared with other deep learning architectures and an elaborative discussion on the same is also mentioned in this paper. We have validated the classification of lung nodules on LUNA dataset and have achieved an accuracy, specificity, sensitivity, area under receiver operating characteristic curve score of 94.7%, 90.41%, 96.68%, 0.98 respectively using fractalnet architecture, which is a substantial improvement over previous works in literature.

NoduleNet: A Lung Nodule Classification Using Deep Learning

Background: Accurate detection and classification of lung nodules at an early stage can help physicians to improve the treatment outcomes of lung cancer. Several lung nodule classifications using deep learning have been proposed but they are lag of external validation to Thai patient data.&#x0D; Objective: To propose a deep learning model called NoduleNet for lung nodule classification and perform internal and external validation of the proposed model.&#x0D; Methods: Two datasets were performed; internal validation using LUNA16 (the public lung CT database), and external validation using ChestRama (37 chest CT scans retrospectively identified from the CT database of Ramathibodi Hospital between 2017 and 2019). The NoduleNet was built on top of pretrained architecture, VGG16, and VGG19 with customization.&#x0D; Results: The NoduleNet showed impressive results in nodule classification. The best model achieved accuracy of 0.95 (0.94 - 0.96), sensitivity of 0.84 (0.82 - 0.86), and specificity of 0.97 (0.97 - 0.98) for internal validation, where the external validation results was accuracy of 0.95 (0.87 - 1.00), sensitivity of 0.91 (0.82 - 1.00), and specificity of 1.00 (1.00 - 1.00). There were 3 misclassified samples in external validation which are all false-negative.&#x0D; Conclusions: The NoduleNet is able to generalize from non-Thai patient data to Thai patient data. It could be further improved by taking sequence of images into account, integrating with an automatic nodule detection algorithm, and adding more nodule types.

MD-NDNet: a multi-dimensional convolutional neural network for false-positive reduction in pulmonary nodule detection

Pulmonary nodule false-positive reduction is of great significance for automated nodule detection in clinical diagnosis of low-dose computed tomography (LDCT) lung cancer screening. Due to individual intra-nodule variations and visual similarities between true nodules and false positives as soft tissues in LDCT images, the current clinical practices remain subject to shortcomings of potential high-risk and time-consumption issues. In this paper, we propose a multi-dimensional nodule detection network (MD-NDNet) for automatic nodule false-positive reduction using deep convolutional neural network (DCNNs). The underlying method collaboratively integrates multi-dimensional nodule information to complementarily and comprehensively extract nodule inter-plane volumetric correlation features using three-dimensional CNNs (3D CNNs) and spatial nodule correlation features from sagittal, coronal, and axial planes using two-dimensional CNNs (2D CNNs) with attention module. To incorporate different sizes and shapes of nodule candidates, a multi-scale ensemble strategy is employed for probability aggregation with weights. The proposed method is evaluated on the LUNA16 challenge dataset in ISBI 2016 with ten-fold cross-validation. Experiment results show that the proposed framework achieves classification performance with a CPM score of 0.9008. All of these indicate that our method enables an efficient, accurate and reliable pulmonary nodule detection for clinical diagnosis.

An adaptive morphology based segmentation technique for lung nodule detection in thoracic CT image

Lung cancer is one of the most life-threatening cancers mostly indicated by the presence of nodules in the lung. Doctors and radiological experts use High-Resolution Computed Tomography (HRCT) images for nodule detection and further decision making from visual inspection. Manual detection of lung nodules is a time-consuming process. Therefore, Computer-aided detection (CADe) systems have been developed for accurate nodule detection and segmentation. CADe-based systems assist radiologists to detect lung nodules with greater confidence and a lesser amount of time and have a significant impact on the accurate, uniform, and early-stage diagnosis of lung cancer. In this research work, an adaptive morphology-based segmentation technique (AMST) has been introduced by designing an adaptive morphological filter for improved segmentation of the lung nodule region. The adaptive morphological filter detects candidate nodule regions by employing adaptive structuring element (ASE) and at the same time improves nodule detection accuracy by reducing false positives (FPs) from the Computed Tomography (CT) slices. The detected nodule candidate regions are then processed for feature extraction. In this study, morphological, texture and intensity-based features have been used with support vector machine (SVM) classifier for lung nodule detection. The performance of the proposed framework has been evaluated by incorporating a 10-fold cross-validation technique on Lung Image Database Consortium-Image Database Resource Initiative (LIDC/IDRI) dataset and on a private dataset, collected from a consultant radiologist. It has been observed that the proposed automated computer-aided detection system has achieved overall classification performance indices with 94.88% sensitivity, 93.45% specificity and 94.27% detection accuracy with 1.8 FPs/scan on LIDC/IDRI dataset and 91.43% sensitivity, 90.45% specificity, 92.83% accuracy with 3.2 FPs/scan on a private dataset. The results show that the proposed CADe system presented in this paper outperforms the other state-of-the-art methods for automatic nodule detection from the HRCT image.

Automatic Detection of Pulmonary Nodules using Three-dimensional Chain Coding and Optimized Random Forest

The detection of pulmonary nodules on computed tomography scans provides a clue for the early diagnosis of lung cancer. Manual detection mandates a heavy radiological workload as it identifies nodules slice-by-slice. This paper presents a fully automated nodule detection with three significant contributions. First, an automated seeded region growing is designed to segment the lung regions from the tomography scans. Second, a three-dimensional chain code algorithm is implemented to refine the border of the segmented lungs. Lastly, nodules inside the lungs are detected using an optimized random forest classifier. The experiments for our proposed detection are conducted using 888 scans from a public dataset, and achieves a favorable result of 93.11% accuracy, 94.86% sensitivity, and 91.37% specificity, with only 0.0863 false positives per exam.