Research on classification of benign and malignant lung nodules based on three-dimensional multi-view squeeze-and-excitation convolutional neural network / 生物医学工程学杂志

Lung cancer is the most threatening tumor disease to human health. Early detection is crucial to improve the survival rate and recovery rate of lung cancer patients. Existing methods use the two-dimensional multi-view framework to learn lung nodules features and simply integrate multi-view features to achieve the classification of benign and malignant lung nodules. However, these methods suffer from the problems of not capturing the spatial features effectively and ignoring the variability of multi-views. Therefore, this paper proposes a three-dimensional (3D) multi-view convolutional neural network (MVCNN) framework. To further solve the problem of different views in the multi-view model, a 3D multi-view squeeze-and-excitation convolution neural network (MVSECNN) model is constructed by introducing the squeeze-and-excitation (SE) module in the feature fusion stage. Finally, statistical methods are used to analyze model predictions and doctor annotations. In the independent test set, the classification accuracy and sensitivity of the model were 96.04% and 98.59% respectively, which were higher than other state-of-the-art methods. The consistency score between the predictions of the model and the pathological diagnosis results was 0.948, which is significantly higher than that between the doctor annotations and the pathological diagnosis results. The methods presented in this paper can effectively learn the spatial heterogeneity of lung nodules and solve the problem of multi-view differences. At the same time, the classification of benign and malignant lung nodules can be achieved, which is of great significance for assisting doctors in clinical diagnosis.

Identification of pulmonary nodule location in three dimensional images and its clinical significance / 中国胸心血管外科临床杂志

@#Objective    To analyze the difference of location identification of pulmonary nodules in two dimensional (2D) and three dimensional (3D) images, and to discuss the identification methods and clinical significance of pulmonary nodules location in 3D space. Methods    The clinical data of 105 patients undergoing sublobectomy in the Department of Thoracic Surgery, the First Affiliated Hospital with Nanjing Medical University from December 2018 to December 2019 were analyzed retrospectively. There were 28 males and 77 females, with an average age of 57.21±13.19 years. The nodule location was determined by traditional 2D method and 3D depth ratio method respectively, and the differences were compared. Results    A total of 30 nodules had different position identification between the two methods, among which 25 nodules in the inner or middle zone of 2D image were located in the peripheral region of 3D image. The overall differences between the two methods were statistically significant (P<0.05). The diagnostic consistency rates of two methods were 66.67% in the right upper lung, 83.33% in the right middle lung, 73.68% in the right lower lung, 75.76% in the left upper lung, and 64.71% in the left lower lung. In each lung lobe, the difference between the two methods in the right upper lung (P=0.014) and the left upper lung (P=0.019) was statistically significant, while in the right middle lung (P=1.000), right lower lung (P=0.460) and left lower lung (P=0.162) were not statistically significant. Conclusion    The 3D position definition of lung nodules based on depth ratio is more accurate than the traditional 2D definition, which is helpful for preoperative planning of sublobectomy.

Application of 3d-printed coplanar template combined with fixed needle technique in percutaneous accurate biopsy of small pulmonary nodules / 中国组织工程研究

BACKGROUND: Percutaneous lung biopsy is an important method to clarify the nature of lung nodules. However, the lungs are more active due to the presence of respiratory motion. Percutaneous lung biopsy, especially for small lung nodules, is difficult. OBJECTIVE: To introduce the application of 3D-printed coplanar template combined with fixed needle technique in percutaneous biopsy of small pulmonary nodules. METHODS: A total of 24 patients who had percutaneous lung biopsy in the Second Affiliated Hospital of Xuzhou Medical University from July 2018 to April 2019 were enrolled. Imaging examination indicated small pulmonary nodules with a nodule diameter of 8-30 mm in all the patients. According to the probability of malignancy, all tumors were of the middle and high risk grade, and there were indications for percutaneous lung biopsy. All the patients were randomized into two groups (n=12 per group): the control group underwent free hand biopsy, and the observation group underwent percutaneous lung biopsy guided by 3D printed coplanar template combined with fixed needle. The number of puncture needle adjustments, number of CT scans, positive rate of specimens, and incidence of complications were recorded and compared between the two groups. Approval for this trial was obtained from the Ethics Committee of the Second Affiliated Hospital of Xuzhou Medical University. RESULTS AND CONCLUSION: The number of puncture needle adjustments, the number of CT scans and the incidence of pneumothorax during the operation were significantly lower in the observation group than the control group (P 0.05). These findings indicate that the 3D-printed coplanar template combined with fixed needle technique can relatively fix the target lesion, reduce the number of needle adjustments and number of CT scans, reduce iatrogenic radiation, and reduce the incidence of complications, especially pneumothorax.

Valor del PET/CT semicuantitativo (SUVmax) en el estudio de nódulo pulmonar sólido / Value of Semiquantitative PET/CT (SUVmax) in study of solid lung nodule

OBJETIVO: Determinar el rendimiento diagnóstico del PET/CT en el estudio de nódulo pulmonar (NP) utilizando SUVmax. MÉTODO: Se revisó la base de datos de PET/CT, seleccionando aquellos solicitados para estudio de NP sólido. Se incluyeron sólo aquellos NP confirmados como malignos o benignos. Se excluyó NP subsólidos, masas pulmonares (> 3 cm), y pacientes con metástasis conocidas. Se midió SUVmax de las lesiones, determinando mejores valores de corte para malignidad y benignidad. RESULTADOS: De los 140 NP estudiados, el 60% (84/140) fueron confirmados como malignos y el 40% como benignos (100% y 59,6% de confirmación histológica, respectivamente). Un SUVmax ≤ 1,0 mostró sensibilidad 98,8%, valor predictivo negativo (VPN) 96,2%, y Likelihood ratio negativo (LR -) 0,027. Un SUVmax ≤ 2,5 no fue capaz de asegurar razonablemente benignidad con VPN 69,4%, y LR - 0,295. Valores de SUV > 2,5 y 5,0 se asociaron a malignidad en 83% y 93% de los casos, respectivamente (LR+ 3,333 y 8,889). CONCLUSIÓN: El PET/CT presenta alto rendimiento diagnóstico en estimar la naturaleza de un NP Un valor de SUVmax ≤ 1 es altamente predictivo de benignidad, y un valor de SUVmax ≥ 2,5 de malignidad. Valores entre 1,0 y 2,5 no permiten caracterizar eficientemente los NP.

AIM: To establish the diagnostic accuracy of PET/CT in study of solid lung nodule (LN) using SUVmax index. METHOD: We revised PET/CT data base, selecting those scans asked to evaluate a solid LN. Only confirmed malign o benign LN were included. Subsolid LN, lung masses (> 3 cm), and known or suspected lung metastases were excluded. SUVmax was measured in each LN, and best cutoff for malignant and benign lesion was calculated. RESULTS: Of the whole group of 140 LN, 60% were confirmed as malignant, and 40% as benign (100% and 59,6% of histological confirmation, respectively). SUVmax ≤ 1,0 showed sensibility of 98,8%, negative predictive value (NPV) of 96,2%, and negative likelihood ratio (LR —) of 0,027. SUVmax ≤ 2,5 was not able to guarantee reasonably benign nature of LN, showing NPV of 69,4% and LR - of 0,295. SUVmax > 2,5 and > 5,0 was associated to malign lesion in 83% and 93% of cases, respectively (LR + of 3,333 and 8,889). CONCLUSION: PET/CT shows high accuracy estimating the nature of solid LN. SUVmax ≤ 1,0 is highly predictive of benignity, and SUVmax ≥ 2,5 is highly predictive of malignancy. SUVmax values between 1,0 and 2,5 were not able to characterize efficiently LN.

Comparison of Short-term Outcomes of Lung Segmentectomy by Robotic-assisted and Video-assisted Thoracoscopic Surgery / 中国肺癌杂志

BACKGROUND@#Lung segmentectomy is increasingly used to resect lung nodules. Video-assisted thoracic surgery (VATS) is widely chosen to performing lung segmentectomy, while robotic assisted thoracoscopic (RATS) was also one useful and practical method. There article was intended to compared the short-time outcomes of RATS and VATS in lung segmentectomy.@*METHODS@#The patients with lung nodules underwent segmentectomy by either RATS or VATS from January 2016 to April 2017 were studied. Baseline characteristics and short-time outcomes (dissected lymph nodes, postoperative duration of drainage, postoperative hospital stay, incidence of pro-longed air leak, atrial fibrillation and pneumonia) were compared.@*RESULTS@#166 patients were included in this study: 81 patients underwent RATS segmentectomy while 85 underwent VATS segmentectomy. The number of lymph nodes dissected in RATS group was more than in VATS group ［(13.07±5.08) vs (10.81±5.74), P=0.010］. The incidence of some postoperative complications such as pro-longed air leak, atrial fibrillation was not significant different between the two approaches.@*CONCLUSIONS@#Compared with VATS, RATS has similar safety and operability, and the number of lymphadenectomy is significantly more than that of VATS.

The effectiveness of deep learning techniques in differentiating benign and malignant lung nodules on target CT scans / 中华放射学杂志

Objective@#To evaluate the effectiveness of deep learning model trained on routine CT scans when identity the malignant and benign lung nodule on target CT scans dataset.@*Methods@#This retrospective study enrolled 923 patients with lung nodules found by chest CT scan in Shanghai Chest Hospital from January 2016 to December 2018. A total of 969 nodules with pathological report were analyzed. The deep learning based pulmonary malignant prediction method in a fine-grained classification manner was used to make the prediction, and the AUC (the area under the curve), accuracy, sensitivity and specificity of routine CT scans and target CT scans were compared, and Delong test and IDI (Integrated Discrimination Improvement) were employed to provide statistical results. Furthermore, statistical methods were used to investigate the differences between the benign and malignant classification of nodules on routine CT and on target CT.@*Results@#In the benign and malignant discrimination task, AUC, accuracy, sensitivity and specificity on the routine scans were 0.81, 82.0%, 86.0% and 56.6% respectively, while the AUC, accuracy, sensitivity and specificity on the target scans were 0.84, 85.0%, 88.8% and 60.5% respectively. The IDI was 0.056 (Z test, P<0.05), and there was statistically significant difference in ROC (Delong test, P=0.01).@*Conclusions@#The deep learning model trained on the data set of routine CT scans can achieve better diagnostic efficiency in target CT scans data.

Clinical Validation of a Protein Biomarker Panel for Non-Small Cell Lung Cancer

We validated the diagnostic performance of a previously developed blood-based 7-protein biomarker panel, AptoDetect™-Lung (Aptamer Sciences Inc., Pohang, Korea) using modified aptamer-based proteomic technology for lung cancer detection. Non-small cell lung cancer (NSCLC), 200 patients and benign nodule controls, 200 participants were enrolled. In a high-risk population corresponding to ≥ 55 years of age and ≥ 30 pack-years, the diagnostic performance was improved, showing 73.3% sensitivity and 90.5% specificity with an area under the curve of 0.88. AptoDetect™-Lung (Aptamer Sciences Inc.) offers the best validated performance to discriminate NSCLC from benign nodule controls in a high-risk population and could play a complementary role in lung cancer screening.

Effect of Hybrid Kernel and Iterative Reconstruction on Objective and Subjective Analysis of Lung Nodule Calcification in Low-Dose Chest CT

OBJECTIVE: To evaluate the differences in subjective calcification detection rates and objective calcium volumes in lung nodules according to different reconstruction methods using hybrid kernel (FC13-H) and iterative reconstruction (IR). MATERIALS AND METHODS: Overall, 35 patients with small (< 4 mm) calcified pulmonary nodules on chest CT were included. Raw data were reconstructed using filtered back projection (FBP) or IR algorithm (AIDR-3D; Canon Medical Systems Corporation), with three types of reconstruction kernel: conventional lung kernel (FC55), FC13-H and conventional soft tissue kernel (FC13). The calcium volumes of pulmonary nodules were quantified using the modified Agatston scoring method. Two radiologists independently interpreted the role of each nodule calcification on the six types of reconstructed images (FC55/FBP, FC55/AIDR-3D, FC13-H/FBP, FC13-H/AIDR-3D, FC13/FBP, and FC13/AIDR-3D). RESULTS: Seventy-eight calcified nodules detected on FC55/FBP images were regarded as reference standards. The calcium detection rates of FC55/AIDR-3D, FC13-H/FBP, FC13-H/AIDR-3D, FC13/FBP, and FC13/AIDR-3D protocols were 80.7%, 15.4%, 6.4%, 52.6%, and 28.2%, respectively, and FC13-H/AIDR-3D showed the smallest calcium detection rate. The calcium volume varied significantly with reconstruction protocols and FC13/AIDR-3D showed the smallest calcium volume (0.04 ± 0.22 mm³), followed by FC13-H/AIDR-3D. CONCLUSION: Hybrid kernel and IR influence subjective detection and objective measurement of calcium in lung nodules, particularly when both techniques (FC13-H/AIDR-3D) are combined.

Centralized Lung Nodule Management at A Veterans Hospital Using A Multidisciplinary Lung Nodule Evaluation Team (LNET) / 中国肺癌杂志

BACKGROUND@#Lung nodules are frequently identified on imaging studies and can represent early lung cancers. We instituted the Lung Nodule Evaluation Team (LNET) to optimize management of these nodules by a lung specialist physician. All lung nodules identified by a radiologist prompted a direct consultation to this service. We report our initial experience with this process.@*METHODS@#This is a retrospective review of patients with lung nodules at a single institution from 2008 to 2015. Since October 2014, lung nodules >3 mm identified on computed tomography (CT) scanning of the chest generate an automatic consult to LNET from the radiology service. Demographic, nodule and follow up data was entered into a surveillance database and summarized.@*RESULTS@#There were 1,873 patients identified in the database. Of these, 900 patients were undergoing active surveillance. Consults increased from 5.5 to 93 per month after the start of the new consult program. Lung nodules were identified on 64% of chest CT scans. Prior to the direct radiology consult the average size of a nodule was 1.7 cm and 0.7 cm after. The overall time from initial nodule imaging to initiating a management plan by a thoracic specialist physician was 3.7 days.@*CONCLUSIONS@#Assessment of lung nodules by a specialist physician is important to ensure appropriate long term management and optimize utilization of diagnostic interventions. A direct radiology consult to a specialized team of chest physicians decreased the time in initiating a management plan, identified smaller nodules and may lead to a more judicious use of health care resources in the management of lung nodules.

Early and Accurate Model of Malignant Lung Nodule Detection System with Less False Positives

ABSTRACT The objective of this work is to identify the malignant lung nodules accurately and early with less false positives. 'Nodule' is the 3mm to 30mm diameter size tissue clusters present inside the lung parenchyma region. Segmenting such a small nodules from consecutive CT scan slices are a challenging task. In our work Auto-seed clustering based segmentation technique is used to segment all the possible nodule candidates. Efficient shape and texture features (2D and 3D) were computed to eliminate the false nodule candidates. The change in centroid position of nodule candidates from consecutive slices was used as a measure to remove the vessels. The two-stage classifier is used in this work to classify the malignant and benign nodules. First stage rule-based classifier producing 100 % sensitivity, but with high false positive of 12.5 per patient scan. The BPN based ANN classifier is used as the second-stage classifier which reduces a false positive to 2.26 per patient scan with a reasonable sensitivity of 88.8%. The Rate of Nodule Growth (RNG) was computed in our work to measure the nodules growth between the two scans of the same patient taken at different time interval. Finally, the nodule growth predictive measure was modeled through the features such as compactness (CO), mass deficit (MD), mass excess (ME) and isotropic factor(IF). The developed model results show that the nodules which have low CO, low IF, high MD and high ME values might have the potential to grow in future.

Research progress on the intraoperative localization of pulmonary ground-glass nodules / 中国胸心血管外科临床杂志

@#With the wide utilization of high-resolution computed tomography (HRCT) in the lung cancer screening, patients detected with pulmonary ground-glass nodules (GGNs) have increased over time and account for a large proportion of all thoracic diseases. Because of its less invasiveness and fast recovery, video-assisted thoracoscopic surgery (VATS) is currently the first choice of surgical approach to lung nodule resection. However, GGNs are usually difficult to recognize during VATS, and failure of nodule localization would result in conversion to thoracotomy or extended lung resection. In order to cope with this problem, a series of approaches for pulmonary nodule localization have developed in the last few years. This article aims to summarize the reported methods of lung nodule localization and analyze its corresponding pros and cons, in order to help thoracic surgeons to choose appropriate localization method in different clinical conditions.

A preliminary investigation on pulmonary subsolid nodule detection using deep learning methods from chest X-rays / 中华放射学杂志

Objective To evaluate the effectiveness of deep learning methods to detect subsolid nodules from chest X-ray images.Methods The building,training,and testing of the deep learning model were performed using the research platform developed by Infervision,China.The training dataset consisted of 1 965 chest X-ray images, which contained 85 labeled subsolid nodules and 1 880 solid nodules. Eighty-five subsolid nodules were confirmed by corresponding CT exams. We labeled each X-ray image using the corresponding reconstructed coronal slice from the CT exam as the gold standard,and trained the deep learning model using alternate training.After the training,the model was tested on a different dataset containing 56 subsolid nodules,which were also confirmed by corresponding coronal slices from CT exams. The model results were compared with an experienced radiologist in terms of sensitivity,specificity,and test time. Results Out of the testing dataset that contained 56 subsolid nodules, the deep learning model marked 72 nodules,which consisted of 39 true positives(TP)and 33 false positives(FP).The model took 17 seconds.The human radiologist marked 39 nodules,with 31 TP and 8 FP.The radiologist took 50 minutes and 24 seconds. Conclusions Subsolid nodules are prone to mis-diagnosis by human radiologists. The proposed deep learning model was able to effectively identify subsolid nodules from X-ray images.

Taxonomic indexes for differentiating malignancy of lung nodules on CT images

Abstract Introduction Lung cancer remains the leading cause of cancer mortality worldwide, with one of the lowest survival rates after diagnosis. Therefore, early detection greatly increases the chances of improving patient survival. Methods This study proposes a method for diagnosis of lung nodules in benign and malignant tumors based on image processing and pattern recognition techniques. Taxonomic indexes and phylogenetic trees were used as texture descriptors, and a Support Vector Machine was used for classification. Results The proposed method shows promising results for accurate diagnosis of benign and malignant lung tumors, achieving an accuracy of 88.44%, sensitivity of 84.22%, specificity of 90.06% and area under the ROC curve of 0.8714. Conclusion The results demonstrate the promising performance of texture extraction techniques by means of taxonomic indexes combined with phylogenetic trees. The proposed method achieves results comparable to those previously published.

CT-guided preoperative hookwire localization of lung nodule in video-assisted thoracic surgery / 中国肿瘤临床

Objective:To explore the feasibility and safety of CT-guided hookwire localization of small lung nodule in video-as-sisted thoracic surgery. Methods: Preoperative localization of small lung nodule was performed using the CT-guided hookwire tech-nique, followed by video-assisted thoracic surgery in the wedge resection. The next mode of operation depends on the results of frozen biopsy. Results:Preoperative localization with CT-guided hookwire was performed in 34 patients between February 2012 and March 2014. The diameter of lung nodule ranged from 5 mm to 22 mm. CT-guided hookwire localization was successful in all patients, with a median positioning time of 23 min. Puncture needles were detached from two of the total patients during the surgery, and three other pa-tients showed pneumothorax by CT scan after localization. Conclusion:Preoperative hookwire localization of small lung nodule is an accurate and safe approach to improve the rate of wedge resection in video-assisted thoracic surgery.

A Computer-Aided Diagnosis for Evaluating Lung Nodules on Chest CT: the Current Status and Perspective

As the detection and characterization of lung nodules are of paramount importance in thoracic radiology, various tools for making a computer-aided diagnosis (CAD) have been developed to improve the diagnostic performance of radiologists in clinical practice. Numerous studies over the years have shown that the CAD system can effectively help readers identify more nodules. Moreover, nodule malignancy and the response of malignant lung tumors to treatment can also be assessed using nodule volumetry. CAD also has the potential to objectively analyze the morphology of nodules and enhance the workflow during the assessment of follow-up studies. Therefore, understanding the current status and limitations of CAD for evaluating lung nodules is essential to effectively apply CAD in clinical practice.

Benign metastasizing leiomyoma.

Benign metastasizing leiomyoma (BML) is a rare condition, affecting predominantly reproductive-age females with uterine leiomyomata and is most often associated with multiple benign-appearing smooth muscle tumors in lungs. We report herein a case of a 38-year-old woman who presented with multiple uterine fibroids for which hysterectomy was carried out on her. Postoperatively, she developed left-sided pleural effusion. Computed chest tomography (CT) scan revealed multiple nodules in both lungs and pleurae. Histopathology of one of the pleura-based nodules revealed a neoplasm composed of interlacing fascicles of spindle cells with uniform nuclei. The tumor cells were positive for alpha-smooth muscle actin and negative for CD34 immunohistochemical stain.

Technological progress of computer-aided detection and diagnosis of lung nodule based on CT image analysis / 国际生物医学工程杂志

Lung nodules are one of the most common pathological changes, thus early detection of lung nodule is very important for the diagnosis medical treatment of lung eancer. In recent years, as the application of multi-slice spiral CT(MSCT), high-resolution CT(HRCT) and low-dose chest CTCLDCT), computer-aided diagnosis (CAD) system will be more essential and more important. Since CAD system can improve the working efficiency of doctors and provide service to more patients, has become the research hotspot and achievement has been made in relevant area internationally recently. This review summarizes the basic methods and applieations of computer-aided detection and diagnosis of lung nodule based on CT image.

Auto-segmentation method for lung parenchyma of CT images / 中国医学影像技术

Objective To establish a new automatic lung segmentation method in order to deal with the omission of pleural nodules and pulmonary vessels. Methods Lung parenchyma were extracted from chest CT images with the inversed operation of 2D region growing and connected area classification, then the contours and locating the contour points were traced with scan line searching. Finally, the parameters of lung contour points were analyzed to locate the contours distorted by nodules, and curve spline was used to correct distorted contours. Results The experimental results of many sets of CT images verified that the technique proposed was effective. The comparison with other contour correction algorithm verified that line searching contour correction was superior. Conclusion The proposed algorithm includes tumors in the segment results, and confirms the integrality, veracity, real-time quality of this auto-segmentation method.

A Case of Double Primary Lung Cancer that was Diagnosed by Percutaneous Localization with using a Hook Wire

With the progress of computed tomography (CT), the detection of small pulmonary nodules has been increased. The conventional diagnostic modalities for tissue confirmation, such as bronchoscopic biopsy or transthoracic needle biopsy, may not be successful in some cases. Too small a nodule or the nodules located far from the pleural surface can be marked and localized with device preoperatively and then this tissue can be obtained surgically. CT-guided hook wire fixation is useful in marking pulmonary nodules and there are few complications with this procedure. We report here on a case of double primary lung cancer that was diagnosed by percutaneous localization with using a hook wire

A New Method of Measuring the Amount of Soft Tissue in Pulmonary Ground-Glass Opacity Nodules: a Phantom Study

OBJECTIVE: To devise a new method to measure the amount of soft tissue in pulmonary ground-glass opacity nodules, and to compare the use of this method with a previous volumetric measurement method by use of a phantom study. MATERIALS AND METHODS: Phantom nodules were prepared with material from fixed normal swine lung. Forty nodules, each with a diameter of 10 mm, were made with a variable mean attenuation. The reference-standard amount of soft tissue in the nodules was obtained by dividing the weight by the specific gravity. The imaging data on the phantom nodules were acquired with the use of a 16-channel multidetector CT scanner. The CT-measured amount of soft tissue of the nodules was calculated as follows: soft tissue amount = volume x (1 + mean attenuation value / 1,000). The relative percentage error (RPE) between the CT-measured amount of the soft tissue and the reference-standard amount of the soft tissue was also measured. The RPEs determined with use of the new method were compared with the RPEs determined with the current volumetric measurement method by the use of the paired t test. RESULTS: The CT-measured amount of soft tissue showed a strong correlation with the reference-standard amount of soft tissue (R(2) = 0.996, p < 0.01). The mean RPE of the CT-measured amount of soft tissue in the nodules was -7.79 +/- 1.88%. The mean RPE of the CT-measured volume was 114.78 +/- 51.02%, which was significantly greater than the RPE of the CT-measured amount of soft tissue (p < 0.01). CONCLUSION: The amount of soft tissue measured by the use of CT reflects the reference-standard amount of soft tissue in the ground-glass opacity nodules much more accurately than does the use of the CT-measured volume.