Multi-View Auxiliary Diagnosis Algorithm for Lung Nodules

Lung is an important organ of human body.More and more people are suffering from lung diseases due to air pollution.These diseases are usually highly infectious.Such as lung tuberculosis,novel coronavirus COVID-19,etc.Lung nodule is a kind of high-density globular lesion in the lung.Physicians need to spend a lot of time and energy to observe the computed tomography image sequences to make a diagnosis,which is inefficient.For this reason,the use of computer-assisted diagnosis of lung nodules has become the current main trend.In the process of computer-aided diagnosis,how to reduce the false positive rate while ensuring a low missed detection rate is a difficulty and focus of current research.To solve this problem,we propose a three-dimensional optimization model to achieve the extraction of suspected regions,improve the traditional deep belief network,and to modify the dispersion matrix between classes.We construct a multi-view model,fuse local three-dimensional information into two-dimensional images,and thereby to reduce the complexity of the algorithm.And alleviate the problem of unbalanced training caused by only a small number of positive samples.Experiments show that the false positive rate of the algorithm proposed in this paper is as low as 12%,which is in line with clinical application standards.

Clinical and Pathological Research Status of Multiple Pulmonary Nodules

With the changes in disease spectrum and the popularization of screening of low-dose spiral CT (CT) in the chest, more and more pulmonary nodules have been detected, most of which are bipulmonary multiple nodules. The existence of multiple pulmonary nodules means that it may be a pathological state of benign and malignant co-existence. The origin and evolution of pulmonary nodules in different histopathological states have a great impact on the choice of treatment methods. In recent years, the rise of immunotherapy has brought a breakthrough in the treatment of refractory lung cancer. However, some patients are still ineffective in immunotherapy, which may be related to the immune microenvironment where nodules are proportioned in different components in different pathological states. This review article mainly predicts the development process of nodules by analyzing the origin of multiple pulmonary nodules and the immune microenvironment of nodules in different pathological conditions, so as to provide guidance for clinical treatment.

Medical Sign Recognition of Lung Nodules Based on Image Retrieval with Semantic Features and Supervised Hashing

Sign recognition is important for identifying benign and malignant nodules. This paper proposes a new sign recognition method based on image retrieval for lung nodules. First, we construct a deep learning framework to extract semantic features that can effectively represent sign information. Second, we translate the high-dimensional image features into compact binary codes with principal component analysis （PCA） and supervised hashing. Third, we retrieve similar lung nodule images with the presented adaptive-weighted similarity calculation method. Finally, we recognize nodule signs from the retrieval results, which can also provide decision support for diagnosis of lung lesions. The proposed method is validated on the publicly available databases： lung image database consortium and image database resource initiative （LIDC-IDRI） and lung computed tomography （CT） imaging signs （LISS）. The experimental results demonstrate our retrieval method substantially improves retrieval performance compared with those using traditional Hamming distance, and the retrieval precision can achieve 87.29%when the length of hash code is 48 bits. The entire recognition rate on the basis of the retrieval results can achieve 93.52%. Moreover, our method is also effective for real-life diagnosis data.

CTCs Detection and Whole-exome Sequencing Might Be Used to Differentiate Benign and Malignant Pulmonary Nodules

Background and objective Low-density computed tomography(LDCT)improved early lung cancer diagnosis but introduces an excess of false-positive pulmonary nodules data.Hence,accurate diagnosis of early-stage lung cancer remains challenging.The purpose of the study was to assess the feasibility of using circulating tumour cells(CTCs)to differentiate malignant from benign pulmonary nodules.Materials and methods 122 patients with suspected malignant pulmonary nodules detected on chest CT in preparation for surgery were prospectively recruited.Peripheral blood samples were collected before surgery,and CTCs were identified upon isolation by size of epithelial tumour cells and morphological analysis.Laser capture microdissection,MALBAC amplification,and whole-exome sequencing were performed on 8 samples.The diagnostic efficacy of CTCs counting,and the genomic variation profile of benign and malignant CTCs samples were analysed.Results Using 2.5 cells/5 m L as the cut-off value,the area under the receiver operating characteristic curve was of 0.651(95%confidence interval:0.538-0.764),with a sensitivity and specificity of 0.526 and 0.800,respectively,and positive and negative predictive values of 91.1%and 30.3%,respectively.Distinct sequence variations differences in DNA damage repair-related and driver genes were observed in benign and malignant samples.TP53 mutations were identified in CTCs of four malignant cases;in particular,g.7578115T>C,g.7578645C>T,and g.7579472G>C were exclusively detected in all four malignant samples.Conclusion CTCs play an ancillary role in the diagnosis of pulmonary nodules.TP53 mutations in CTCs might be used to identify benign and malignant pulmonary nodules.

End-to-End 2D Convolutional Neural Network Architecture for Lung Nodule Identification and Abnormal Detection in Cloud

The extent of the peril associated with cancer can be perceivedfrom the lack of treatment, ineffective early diagnosis techniques, and mostimportantly its fatality rate. Globally, cancer is the second leading cause ofdeath and among over a hundred types of cancer;lung cancer is the secondmost common type of cancer as well as the leading cause of cancer-relateddeaths. Anyhow, an accurate lung cancer diagnosis in a timely manner canelevate the likelihood of survival by a noticeable margin and medical imagingis a prevalent manner of cancer diagnosis since it is easily accessible to peoplearound the globe. Nonetheless, this is not eminently efficacious consideringhuman inspection of medical images can yield a high false positive rate. Ineffectiveand inefficient diagnosis is a crucial reason for such a high mortalityrate for this malady. However, the conspicuous advancements in deep learningand artificial intelligence have stimulated the development of exceedinglyprecise diagnosis systems. The development and performance of these systemsrely prominently on the data that is used to train these systems. A standardproblem witnessed in publicly available medical image datasets is the severeimbalance of data between different classes. This grave imbalance of data canmake a deep learning model biased towards the dominant class and unableto generalize. This study aims to present an end-to-end convolutional neuralnetwork that can accurately differentiate lung nodules from non-nodules andreduce the false positive rate to a bare minimum. To tackle the problem ofdata imbalance, we oversampled the data by transforming available images inthe minority class. The average false positive rate in the proposed method isa mere 1.5 percent. However, the average false negative rate is 31.76 percent.The proposed neural network has 68.66 percent sensitivity and 98.42 percentspecificity.

Attention Based Multi-Patched 3D-CNNs with Hybrid Fusion Architecture for Reducing False Positives during Lung Nodule Detection

In lung nodules there is a huge variation in structural properties like Shape, Surface Texture. Even the spatial properties vary, where they can be found attached to lung walls, blood vessels in complex non-homogenous lung structures. Moreover, the nodules are of small size at their early stage of development. This poses a serious challenge to develop a Computer aided diagnosis (CAD) system with better false positive reduction. Hence, to reduce the false positives per scan and to deal with the challenges mentioned, this paper proposes a set of three diverse 3D Attention based CNN architectures (3D ACNN) whose predictions on given low dose Volumetric Computed Tomography (CT) scans are fused to achieve more effective and reliable results. Attention mechanism is employed to selectively concentrate/weigh more on nodule specific features and less weight age over other irrelevant features. By using this attention based mechanism in CNN unlike traditional methods there was a significant gain in the classification performance. Contextual dependencies are also taken into account by giving three patches of different sizes surrounding the nodule as input to the ACNN architectures. The system is trained and validated using a publicly available LUNA16 dataset in a 10 fold cross validation approach where a competition performance metric (CPM) score of 0.931 is achieved. The experimental results demonstrate that either a single patch or a single architecture in a one-to-one fashion that is adopted in earlier methods cannot achieve a better performance and signifies the necessity of fusing different multi patched architectures. Though the proposed system is mainly designed for pulmonary nodule detection it can be easily extended to classification tasks of any other 3D medical diagnostic computed tomography images where there is a huge variation and uncertainty in classification.

Clinical characteristics and work-up of small to intermediate-sized pulmonary nodules in a Chinese dedicated cancer hospital

Objectives:To evaluate the characteristics and work-up of small to intermediate-sized pulmonary nodules in a Chinese dedicated cancer hospital.Methods:Patients with pulmonary nodules 4–25 mm in diameter detected via computed tomography(CT)in 2013 were consecutively included.The analysis was restricted to patients with a histological nodule diagnosis or a 2-year follow-up period without nodule growth confirming benign disease.Patient information was collected from hospital records.Results:Among the 314 nodules examined in 299 patients,212(67.5%)nodules in 206(68.9%)patients were malignant.Compared to benign nodules,malignant nodules were larger(18.0 mm vs.12.5 mm,P<0.001),more often partly solid(16.0%vs.4.7%,P<0.001)and more often spiculated(72.2%vs.41.2%,P<0.001),with higher density in contrast-enhanced CT(67.0 HU vs.57.5 HU,P=0.015).Final diagnosis was based on surgery in 232 out of 314(73.9%)nodules,166 of which were identified as malignant[30(18.1%)stage III or IV]and 66 as benign.In 36 nodules(11.5%),diagnosis was confirmed by biopsy and the remainder verified based on stability of nodule size at follow-up imaging(n=46,14.6%).Among 65 nodules subjected to gene(EGFR)mutation analyses,28(43.1%)cases(EGFR19 n=13;EGFR21 n=15)were identified as EGFR mutant and 37(56.9%)as EGFR wild-type.Prior to surgery,the majority of patients[n=194(83.6%)]received a contrast-enhanced CT scan for staging of both malignant[n=140(84.3%)]and benign[n=54(81.8%)]nodules.Usage of positron emission tomography(PET)-CT was relatively uncommon[n=38(16.4%)].Conclusions:CT-derived nodule assessment assists in diagnosis of small to intermediate-sized malignant pulmonary nodules.Currently,contrast-enhanced CT is commonly used as the sole diagnostic confirmation technique for pre-surgical staging,often resulting in surgery for late-stage disease and unnecessary surgery in cases of benign nodules.

Simultaneous Paragonimus infection involving the breast and lung:A case report

BACKGROUND Paragonimiasis is a food-borne parasitic infection caused by lung flukes of the genus Paragonimus. Although the most common site of infection is the pleuropulmonary area, the parasite can also reach other parts of the body on its journey from the intestines to the lungs, ending up in locations such as the brain,abdomen, skin, and subcutaneous tissues. Ectopic paragonimiasis is difficult to diagnose due to the rarity of this disease.CASE SUMMARY Here, we report a rare case of simultaneous breast and pulmonary paragonimiasis in a woman presenting painless breast mass and lung nodule with a history of eating raw trout. To confirm the diagnosis, serologic testing and tissue confirmation of the breast mass were performed. The patient was treated with surgical resection of the mass and praziquantel medication.CONCLUSION Ectopic paragonimiasis is difficult to diagnose due to the rarity of this disease.Thus, thorough history-taking and clinical suspicion of parasitic infection are important.

Imaging diagnosis of bronchogenic carcinoma(the forgotten disease)during times of COVID-19 pandemic:Current and future perspectives

Patients with bronchogenic carcinoma comprise a high-risk group for coronavirus disease 2019(COVID-19),pneumonia and related complications.Symptoms of COVID-19 related pulmonary syndrome may be similar to deteriorating symptoms encountered during bronchogenic carcinoma progression.These resemblances add further complexity for imaging assessment of bronchogenic carcinoma.Similarities between clinical and imaging findings can pose a major challenge to clinicians in distinguishing COVID-19 super-infection from evolving bronchogenic carcinoma,as the above-mentioned entities require very different therapeutic approaches.However,the goal of bronchogenic carcinoma management during the pandemic is to minimize the risk of exposing patients to COVID-19,whilst still managing all life-threatening events related to bronchogenic carcinoma.The current pandemic has forced all healthcare stakeholders to prioritize per value resources and reorganize therapeutic strategies for timely management of patients with COVID-19 related pulmonary syndrome.Processing of radiographic and computed tomography images by means of artificial intelligence techniques can facilitate triage of patients.Modified and newer therapeutic strategies for patients with bronchogenic carcinoma have been adopted by oncologists around the world for providing uncompromised care within the accepted standards and new guidelines.

Pulmonary benign metastasizing leiomyoma: A case report and review of the literature

BACKGROUND Pulmonary benign metastatic leiomyoma(PBML),which is very rare,is a type of benign metastatic leiomyoma(BML).Here,we report a case of PBML,finally diagnosed through multidisciplinary team(MDT)discussions,and provide a literature review of the disease.CASE SUMMARY A 55-year old asymptomatic woman was found to have bilateral multiple lung nodules on a chest high-resolution computed tomography(HRCT)scan.Her medical history included total hysterectomy for uterine leiomyoma.The patient was diagnosed with PBML,on the basis of her clinical history,imaging manifestations,and computed tomography(CT)-guided percutaneous lung puncture biopsy,via MDT discussions.As the patient was asymptomatic,she received long-term monitoring without treatment.A follow-up of chest HRCT after 6 mo showed that the PBML lung nodules were stable and there was no progression.CONCLUSION For patients with a medical history of hysterectomy and uterine leiomyoma with lung nodules on chest CT,PBML should be considered during diagnosis based on the clinical history,imaging manifestations,CT-guided percutaneous lung puncture biopsy,and MDT discussions.

Algorithm fusion to improve detection of lung cancer on chest radiographs

Purpose-The purpose of this paper is to show an efficient method for the detection of signs of early lung cancer.Various image processing algorithms are presented for different types of lesions,and a scheme is proposed for the combination of results.Design/methodology/approach-A computer aided detection(CAD)scheme was developed for detection of lung cancer.It enables different lesion enhancer algorithms,sensitive to specific lesion subtypes,to be used simultaneously.Three image processing algorithms are presented for the detection of small nodules,large ones,and infiltrated areas.The outputs are merged,the false detection rate is reduced with four separated support vector machine(SVM)classifiers.The classifier input comes from a feature selection algorithm selecting from various textural and geometric features.A total of 761 images were used for testing,including the database of the Japanese Society of Radiological Technology(JSRT).Findings-The fusion of algorithms reduced false positives on average by 0.6 per image,while the sensitivity remained 80 per cent.On the JSRT database the system managed to find 60.2 per cent of lesions at an average of 2.0 false positives per image.The effect of using different result evaluation criteria was tested and a difference as high as 4 percentage points in sensitivity was measured.The system was compared to other published methods.Originality/value-The study described in the paper proves the usefulness of lesion enhancement decomposition,while proposing a scheme for the fusion of algorithms.Furthermore,a new algorithm is introduced for the detection of infiltrated areas,possible signs of lung cancer,neglected by previous solutions.

Amplification Method of Lung Nodule Data Based on DCGAN Generation Algorithm

Early diagnosis of lung cancer can effectively reduce the mortality of patients.Doctors use low-dose spiral CT to detect lung nodules,which is timeconsuming and prone to omissions.Deep learning has achieved good results in the field of medical image sub-processing,which can reduce the pressure of doctors to a certain extent.However,in the actual lung CT images,the images containing lung nodules account for less than 1%of the total images.The lack of data increases the difficulty of detecting lung nodules by using deep learning methods.This paper proposes an amplification method using deep convolutional anti-generation network(DCGAN)to generate lung nodule data.Compared with different amplification methods,and the effectiveness of this method is confirmed.Experiments can prove that the use of DCGAN to generate data can better solve the problems of high false positive rate and low sensitivity of lung nodule classification than the graphical data amplification mode.Compared with the existing methods,this experimental method greatly improves the accuracy,sensitivity and F1 score of lung nodule detection,and achieves good results of 99.98%,99.15%and 99.55%,respectively.