Classification Of Interstitial Lung Disease Research Articles

Pattern classification of interstitial lung diseases from computed tomography images using a ResNet-based network with a split-transform-merge strategy and split attention.

In patients with interstitial lung disease (ILD), accurate pattern assessment from their computed tomography (CT) images could help track lung abnormalities and evaluate treatment efficacy. Based on excellent image classification performance, convolutional neural networks (CNNs) have been massively investigated for classifying and labeling pathological patterns in the CT images of ILD patients. However, previous studies rarely considered the three-dimensional (3D) structure of the pathological patterns of ILD and used two-dimensional network input. In addition, ResNet-based networks such as SE-ResNet and ResNeXt with high classification performance have not been used for pattern classification of ILD. This study proposed a SE-ResNeXt-SA-18 for classifying pathological patterns of ILD. The SE-ResNeXt-SA-18 integrated the multipath design of the ResNeXt and the feature weighting of the squeeze-and-excitation network with split attention. The classification performance of the SE-ResNeXt-SA-18 was compared with the ResNet-18 and SE-ResNeXt-18. The influence of the input patch size on classification performance was also evaluated. Results show that the classification accuracy was increased with the increase of the patch size. With a 32 × 32 × 16 input, the SE-ResNeXt-SA-18 presented the highest performance with average accuracy, sensitivity, and specificity of 0.991, 0.979, and 0.994. High-weight regions in the class activation maps of the SE-ResNeXt-SA-18 also matched the specific pattern features. In comparison, the performance of the SE-ResNeXt-SA-18 is superior to the previously reported CNNs in classifying the ILD patterns. We concluded that the SE-ResNeXt-SA-18 could help track or monitor the progress of ILD through accuracy pattern classification.

Segmentation and Classification of Interstitial Lung Diseases Based on Hybrid Deep Learning Network Model

Segmentation and Classification of Interstitial Lung Diseases Based on Hybrid Deep Learning Network Model

ИНТЕРСТИЦИАЛЬНЫЕ ЗАБОЛЕВАНИЯ ЛЕГКИХ У ДЕТЕЙ: СОВРЕМЕННАЯ КЛАССИФИКАЦИЯ, АЛГОРИТМ ДИАГНОСТИКИ, ОБЩИЕ ПОДХОДЫ К ТЕРАПИИ

Authors, based on bibliographical research and their own data, provide up-to-date information on interstitial lung diseases (ILD) in children. The classification of ILD in children that is used in the registry of pediatric patients with ILD in the European Union (2022) with a critical analysis is given, and the need for its practical use is substantiated. The algorithm for diagnosing childhood ILD is described in detail starting with the diagnostic criteria for childhood ILD syndrome and possible complications. The algorithm includes step-by-step, sequential steps to diagnose these rare diseases. Taking into account the latest advances in genetics and molecular biology, the main proteins and their encoding genes associated with ILD, types of inheritance and phenotypes are listed as well. Indications and findings during lung biopsy, general approaches to the treatment of ILD in children, including oxygen therapy, drug therapy and other therapy, taking into account the existing consensus of experts, are characterized.

Clinical Characterization and Predictive Factors for Progression in a Cohort of Patients with Interstitial Lung Disease and Features of Autoimmunity: The Need for a Revision of IPAF Classification Criteria.

Background and Objectives: The "interstitial pneumonia with autoimmune features" (IPAF) criteria have been criticized because of the exclusion of usual interstitial pneumonia (UIP) patients with a single clinical or serological feature. To classify these patients, the term UIPAF was proposed. This study aims to describe clinical characteristics and predictive factors for progression of a cohort of interstitial lung disease (ILD) patients with at least one feature of autoimmunity, applying criteria for IPAF, specific connective tissue diseases (CTD), and a definition of UIPAF when possible. Methods: We retrospectively evaluated data on 133 consecutive patients with ILD at onset associated with at least one feature of autoimmunity, referred by pulmonologists to rheumatologists from March 2009 to March 2020. Patients received 33 (16.5-69.5) months of follow-up. Results: Among the 101 ILD patients included, 37 were diagnosed with IPAF, 53 with ILD-onset CTD, and 11 with UIPAF. IPAF patients had a lower prevalence of UIP pattern compared to CTD-ILD and UIPAF patients (10.8% vs. 32.1% vs. 100%, p < 0.01). During the follow-up, 4 IPAF (10.8%) and 2 UIPAF (18.2%) patients evolved into CTD-ILD. IPAF patients presented features not included in IPAF criteria, such as sicca syndrome (8.1%), and were more frequently affected by systemic hypertension (p < 0.01). Over one year, ILD progression (greater extent of fibrosis on HRCT and/or decline in PFTs) was less frequent in the IPAF group compared to CTD-ILD and UIPAF (32.3% vs. 58.8% vs. 72.7, p = 0.02). A UIP pattern and an IPAF predicted a faster (OR: 3.80, p = 0.01) and a slower (OR: 0.28, p = 0.02) ILD progression, respectively. Conclusions: IPAF criteria help identify patients who might develop a CTD-ILD, even though a single clinical or serological feature is respected. Future revisions of IPAF criteria should include sicca syndrome and separate UIP-pattern into a different definition (UIPAF), given its association with a different prognosis, independently from ILD classification.

Influence of thoracic radiology training on classification of interstitial lung diseases

IntroductionInterpretation of high-resolution CT images plays an important role in the diagnosis and management of interstitial lung diseases. However, interreader variation may exist due to varying levels of training and expertise. This study aims to evaluate interreader variation and the role of thoracic radiology training in classifying interstitial lung disease (ILD). MethodsThis is a retrospective study where seven physicians (radiologists, thoracic radiologists, and a pulmonologist) classified the subtypes of ILD of 128 patients from a tertiary referral center, all selected from the Interstitial Lung Disease Registry which consists of patients from November 2014 to January 2021. Each patient was diagnosed with a subtype of interstitial lung disease by a consensus diagnosis from pathology, radiology, and pulmonology. Each reader was provided with only clinical history, only CT images, or both. Reader sensitivity and specificity and interreader agreements using Cohen's κ were calculated. ResultsInterreader agreement based only on clinical history, only on radiologic information, or combination of both was most consistent amongst readers with thoracic radiology training, ranging from fair (Cohen's κ: 0.2–0.46), moderate to almost perfect (Cohen's κ: 0.55–0.92), and moderate to almost perfect (Cohen's κ: 0.53–0.91) respectively. Radiologists with any thoracic training showed both increased sensitivity and specificity for NSIP as compared to other radiologists and the pulmonologist when using only clinical history, only CT information, or combination of both (p < 0.05). ConclusionsReaders with thoracic radiology training showed the least interreader variation and were more sensitive and specific at classifying certain subtypes of ILD. Summary sentenceThoracic radiology training may improve sensitivity and specificity in classifying ILD based on HRCT images and clinical history.

A computer aided diagnosis framework for detection and classification of interstitial lung diseases using computed tomography (CT) images

A computer aided diagnosis framework for detection and classification of interstitial lung diseases using computed tomography (CT) images

Two-Stage Hybrid Approach of Deep Learning Networks for Interstitial Lung Disease Classification.

High-resolution computed tomography (HRCT) images in interstitial lung disease (ILD) screening can help improve healthcare quality. However, most of the earlier ILD classification work involves time-consuming manual identification of the region of interest (ROI) from the lung HRCT image before applying the deep learning classification algorithm. This paper has developed a two-stage hybrid approach of deep learning networks for ILD classification. A conditional generative adversarial network (c-GAN) has segmented the lung part from the HRCT images at the first stage. The c-GAN with multiscale feature extraction module has been used for accurate lung segmentation from the HRCT images with lung abnormalities. At the second stage, a pretrained ResNet50 has been used to extract the features from the segmented lung image for classification into six ILD classes using the support vector machine classifier. The proposed two-stage algorithm takes a whole HRCT as input eliminating the need for extracting the ROI and classifies the given HRCT image into an ILD class. The performance of the proposed two-stage deep learning network-based ILD classifier has improved considerably due to the stage-wise improvement of deep learning algorithm performance.

Immunomodulatory treatment of interstitial lung disease.

Interstitial lung diseases (ILDs) other than idiopathic pulmonary fibrosis (IPF) have an array of immunomodulatory treatment options compared with IPF, due to their inflammatory component. However, there is a relative paucity of guidance on the management of this heterogeneous group of diseases. In ILDs other than IPF, immunosuppression is the cornerstone of therapy, with varying levels of evidence for different immunomodulatory agents and for each specific ILD. Classification of ILDs is important for guiding treatment decisions. Immunomodulatory agents mainly include corticosteroids, mycophenolate mofetil (MMF), azathioprine, methotrexate, cyclophosphamide and rituximab. In this review, the available evidence for single agents in the most common ILDs is first discussed. We then reviewed practical therapeutic approaches in connective tissue disease–related ILD and interstitial pneumonia with autoimmune features, scleroderma-related ILD, vasculitis and dermatomyositis with hypoxemic respiratory failure, idiopathic non-specific interstitial pneumonia, hypersensitivity pneumonitis sarcoidosis, fibrosing organizing pneumonia and eosinophilic pneumonia. The treatment of acute exacerbations of ILD is also discussed. Therapy augmentation in ILD is dictated by the recognition of progression of disease. Criteria for the evaluation of progression of disease are then discussed. Finally, specific protocol and measures to increase patients’ safety are reviewed as well, including general monitoring and serologic surveillance, Pneumocystis jirovecii prophylaxis, patients’ education, genetic testing for azathioprine, MMF serum levels and cyclophosphamide administration protocols. Immunomodulatory therapies are largely successful in the management of ILDs and can be safely managed with the application of specific protocols, precautions and monitoring.

Interstitial lung disease revisited

Interstitial lung disease (ILD), consisting of more than 200 subtypes of diseases, is a large group of heterogeneous diseases characterized by varying degrees of alveolitis and pulmonary parenchymal fibrosis. There are many challenges in its diagnosis and treatment. In recent years, new understanding of the clinical classification, disease behavior, early diagnosis, and treatment of ILD has been obtained. The anti-fibrotic drugs have been expanded successfully from treatment of idiopathic pulmonary fibrosis (IPF) to progressive-fibrosing interstitial lung diseases (PF-ILD). A new international clinical practice guideline categorized hypersensitivity pneumonitis (HP) into two clinical phenotypes, namely nonfibrotic and fibrotic HP. The clinical classification of ILD according to disease behavior and a progressing fibrotic phenotype is of important value for the establishment of treatment strategies for patients with ILD. Interstitial pulmonary abnormalities (ILA) may represent the subclinical stages of different types of ILD, and long-term follow-up of ILA is key to improving the early diagnosis of ILD. Interstitial pneumonia with autoimmune features (IPAF) may represent an early phase or prodromal state of a connective tissue disease (CTD), and patients with IPAF need to be under longitudinal surveillance for evolution to CTD.

Detection, screening, and classification of interstitial lung disease in patients with systemic sclerosis.

Interstitial lung disease (ILD) associates with disease burden and reduced life expectancy in systemic sclerosis (SSc). As ILD afflicts around 50% of SSc patients and is often present from early on, there is rationale for early and universal ILD screening. With the approval of the first SSc-ILD therapy last year, there is an increasing need for ILD classifications to assign the right treatment to the right patient. Here, we discuss recent advances on screening, detection, and classification of SSc-ILD. Although prospective datasets from a nationwide population-based SSc cohort confirm insufficient sensitivity of pulmonary function tests (PFTs) for ILD screening, they provide strong support for lung high resolution computed tomography (HRCT) as the primary tool to detect ILD. Lung ultrasound shows promise as an additional screening tool. Interpreting statements from a new European consensus on SSc-ILD management, we propose an urgent need for integrated SSc classification, grading ILD severity at time of diagnosis and evaluate risk for ILD progression. We discuss advances on potential parameters for such classification, including PFTs, quantitative HRCT analyses, patient-reported outcome measures, functional exercise capacity tests, and soluble biomarkers. Early screening to diagnose ILD is feasible. With new therapies at hand, there is a need for integrated ILD classification including severity grading and risk for progression.

Utility of Bronchoalveolar Lavage and Transbronchial Biopsy in Patients with Interstitial Lung Disease.

Bronchoalveolar lavage and transbronchial biopsy can be a useful tool in the evaluation of interstitial lung disease (ILD), but patient selection for this procedure remains poorly defined. Determining clinical characteristics that help with patient selection for bronchoscopy may improve confidence of ILD classification while limiting potential adverse outcomes associated with surgical lung biopsy. The purpose of this study is to identify factors that were associated with change in multidisciplinary ILD diagnosis (MDD) before and after incorporation of BAL and TBBx data. We conducted a retrospective cohort study of ILD patients at a single center who underwent bronchoscopy in the diagnostic workup of ILD. We performed sequential MDD both pre- and post-bronchoscopy to calculate the frequency of change in diagnosis after incorporating information from BAL and TBBx and identify features associated with change in diagnosis. 245 patients were included in the study. Bronchoscopy led to a change in diagnosis in 58 patients (23.7%). The addition of TBBx to BAL increased diagnostic yield from 21.8 to 34.1% (p = 0.027). Identification of antigen, HRCT scan inconsistent with UIP, and absence of a pre-bronchoscopy diagnosis of CTD-ILD or IPAF were associated with a change in diagnosis after bronchoscopy. Our study suggests clinical features that may assist with patient selection for bronchoscopy. We suggest bronchoscopy in patients with identified antigen or an HRCT that is consistent with a non-IPF diagnosis. Appropriate patient selection for bronchoscopy may improve ILD diagnostic confidence and avoid potential complications from more invasive and higher risk procedures.

Deep Learning of Computed Tomography Virtual Wedge Resection for Prediction of Histologic Usual Interstitial Pneumonitis.

Rationale: The computed tomography (CT) pattern of definite or probable usual interstitial pneumonia (UIP) can be diagnostic of idiopathic pulmonary fibrosis and may obviate the need for invasive surgical biopsy. Few machine-learning studies have investigated the classification of interstitial lung disease (ILD) on CT imaging, but none have used histopathology as a reference standard.Objectives: To predict histopathologic UIP using deep learning of high-resolution computed tomography (HRCT).Methods: Institutional databases were retrospectively searched for consecutive patients with ILD, HRCT, and diagnostic histopathology from 2011 to 2014 (training cohort) and from 2016 to 2017 (testing cohort). A blinded expert radiologist and pulmonologist reviewed all training HRCT scans in consensus and classified HRCT scans based on the 2018 American Thoracic Society/European Respriatory Society/Japanese Respiratory Society/Latin American Thoracic Association diagnostic criteria for idiopathic pulmonary fibrosis. A convolutional neural network (CNN) was built accepting 4 × 4 × 2 cm virtual wedges of peripheral lung on HRCT as input and outputting the UIP histopathologic pattern. The CNN was trained and evaluated on the training cohort using fivefold cross validation and was then tested on the hold-out testing cohort. CNN and human performance were compared in the training cohort. Logistic regression and survival analyses were performed.Results: The CNN was trained on 221 patients (median age 60 yr; interquartile range [IQR], 53-66), including 71 patients (32%) with UIP or probable UIP histopathologic patterns. The CNN was tested on a separate hold-out cohort of 80 patients (median age 66 yr; IQR, 58-69), including 22 patients (27%) with UIP or probable UIP histopathologic patterns. An average of 516 wedges were generated per patient. The percentage of wedges with CNN-predicted UIP yielded a cross validation area under the curve of 74% for histopathological UIP pattern per patient. The optimal cutoff point for classifying patients on the training cohort was 16.5% of virtual lung wedges with CNN-predicted UIP and resulted in sensitivity and specificity of 74% and 58%, respectively, in the testing cohort. CNN-predicted UIP was associated with an increased risk of death or lung transplantation during cross validation (hazard ratio, 1.5; 95% confidence interval, 1.1-2.2; P = 0.03).Conclusions: Virtual lung wedge resection in patients with ILD can be used as an input to a CNN for predicting the histopathologic UIP pattern and transplant-free survival.

Deep Convolutional Neural Networks for Classification of Interstitial Lung Disease

Automated lung tissue characterization of Interstitial Lung Disease is one of the most important aspects of the Computer Aided Disease diagnosis system. The problem remains challenging, even though there has been much research in this area. While deep learning has produced brilliant success in image applications over the past few years, the majority of training is with sub-optimal parameters, requiring unnecessary long training time, setting up hyper parameters. In this paper, we explore the classification of lung tissue pattern affected with interstitial lung disease (ILD) in high resolution computed tomography (HRCT) scans and evaluated different CNN architectures with and without transfer learning. The effect of cyclical learning rates, the hyper-parameters tuning and data augmentation on classification performance are studied using a popular publicly available dataset called MedGift dataset.

Filtering Method for Pre processing Mammogram Images for Breast Cancer Detection

Breast cancer is a stand-out surrounded by the most widely perceived diseases and has a high rate of mortality around the world, significantly risking the health of the females. Among existing all modalities of medical scans, mammography is the most preferred modality for preliminary examination of breast cancer. To assist radiologist, a computer-aided diagnosis (CAD) is enhancing and important medical systems for mammographic lesion analysis. CAD is necessary to provide doctors, to improve detection quality of breast cancer. In mammogram images, micro-calcifications is one of the imperative sign for breast cancer detection. Mammographic medical scan may present unwanted noise and CAD systems are very sensitive to noise. So, pre-processing of medical images for any medical image analysis application like brain tumor detection, breast cancer detection, and interstitial lung disease classification is considered as an important step. The segmentation or classification accuracy is mainly depends upon the significant improved pre-processing process. Thus, in this work, different types of filtering techniques used for noise reduction in medical image processing are analyzed. The qualitative and quantitative results are examined on mini-MIAS mammogram image database. The effectiveness of filtering techniques is compared based on the different quantitative parameters and visual qualities of examined output.

Multidisciplinary approach in the diagnosis of idiopathic nonspecific interstitial pneumonia.

The article provides a modern classification of interstitial lung diseases. The focus is on the poorly studied and difficult to diagnose idiopathic nonspecific interstitial pneumonia, isolated in a separate nosological form and included in the classification only in 2002. The paper presents the features of the clinical, radiological, histological picture of idiopathic nonspecific interstitial pneumonia in comparison with idiopathic pulmonary fibrosis. The presented materials are based on the results of the largest study conducted by the working group of the American Thoracic Society, which shows the importance of the interaction of specialists in the diagnosis of idiopathic nonspecific interstitial pneumonia. A comprehensive assessment of the examination of patients conducted in collaboration with the clinician and radiologist, after their cooperative discussion, in many cases can help to avoid lung biopsy, and the study of morphological material is more often necessary only in difficult diagnostic situations.

Tissue Continues to Be the Issue: Role of Histopathology in the Context of Recent Updates in the Radiologic Classification of Interstitial Lung Diseases

High-resolution computed tomography (HRCT) imaging has an increasingly important role in clinical decision-making in patients with interstitial lung diseases. The recent Fleischner Society white paper on the diagnostic criteria for idiopathic pulmonary fibrosis highlights the advances in our understanding of HRCT imaging in interstitial lung diseases. To discuss the evidence and recommendations outlined in the white paper as it pertains to the radiologic diagnosis of interstitial lung disease, specifically highlighting the current limitations of HRCT in confidently predicting histopathologic findings. The recent Fleischner Society white paper and other studies pertaining to the role of HRCT in predicting histopathology in interstitial lung diseases are reviewed. High-resolution computed tomography is highly predictive of a usual interstitial pneumonia (UIP) pattern on histopathology when the HRCT shows a typical UIP pattern on a "confident" read by the radiologist. A probable UIP pattern is also very predictive of a UIP pattern on histopathology, and histopathologic confirmation is not needed for most patients demonstrating this pattern in the appropriate clinical setting. A UIP pattern may be seen in a substantial proportion of patients with an "indeterminate UIP" pattern on HRCT and in many patients for whom the HRCT suggests an alternative diagnosis; histopathologic confirmation should be considered in patients demonstrating these patterns whenever feasible.

Interstitial lung disease classification using improved DenseNet

Interstitial Lung Disease (ILD) is one of the popular respiratory diseases. The correct diagnosis of ILD is beneficial to improve the effect of treatment for patients. This paper presents an improved DenseNet called small kernel DenseNet (SK-DenseNet) to improve ILD classification performance. According to the characteristics of HRCT features of lung disease, the SK-DenseNet network is more effective to extract high level and small pathological features for ILD classification. Our experiment results show that the proposed SK-DenseNet obtains an outstanding performance (~98.4%),which improves 5% performance compared with DenseNet. A comparative analysis with other CNNs, such as AlexNet, VGGNet, ResNet has also demonstrated that the effectiveness of SK-DenseNet in terms of classifying lung disease patterns is superior than those compared ones. The research has validated that using small convolution kernel is useful to improve the recognition efficiency when feature patterns are small.

Pattern classification of interstitial lung disease in high resolution clinical datasets: A systematic review

Automated tissues characterization helps to diagnosis the various diseases including Interstitial lung diseases (ILD). The various features and the several classifiers are used in categorize the different layers depend on the pattern presented in the image. The different types of diseases may occur in the lungs and some of the diseases happen to leave the scars. These scars can be found in the High Resolution Computed Tomography (HRCT) and have different pattern. The different diseases cause the different pattern in the images and these is classified using the efficient classifier that helps to diagnosis the diseases. In this paper, review for the many researches regarding to the classification of the different pattern from the Computed Tomography (CT) images is presented. The evaluation of the efficiency of the methods in terms of classifier and database used for the research is made. The Deep Convolution Neural Network (CNN) provides the promising classifier efficiency compared to the other researches for different pattern. In general, there are five types of pattern is classified: Healthy, ground glass, honeycomb, Fibrosis, and emphysema.

Impact of Interstitial Lung Disease Classification on the Development of Acute Exacerbation of Interstitial Lung Disease and Prognosis in Patients with Stage III Non-Small-Cell Lung Cancer and Interstitial Lung Disease Treated With Chemoradiotherapy

Introduction: Data on the efficacy and risk of curative-intent chemoradiotherapy in patients with inoperable stage III non-small-cell lung cancer (NSCLC) and interstitial lung disease (ILD) are limited. The aim of this study was to explore the impact of ILD classification on acute exacerbation (AE) of ILD and prognosis in patients with stage III NSCLC and ILD treated with chemoradiotherapy.Materials and methods: We retrospectively reviewed the medical records of patients with stage III NSCLC and ILD treated with curative-intent chemoradiotherapy as the first-line treatment at the Shizuoka Cancer Center between June 2009 and May 2014.Results: Of 37 patients, 17 (46%) developed AE of ILD worse than grade 3 within 1 year after the last irradiation. In univariate analysis, the incidence rate of AE of ILD was lower in patients with a non-usual interstitial pneumonia (UIP) pattern than in those with a UIP pattern. Multivariate analysis showed that ILD classification was significantly associated with the incidence of AE of ILD. The median overall survival (OS) durations in patients with a non-UIP pattern and a UIP pattern were 16.5 and 9.3 months, respectively. In univariate analysis, patients with a non-UIP pattern showed better survival. Multivariate analysis showed that ILD classification was a significant independent prognostic factor.Conclusion: The incidence of AE of ILD was high in patients with stage III NSCLC and ILD treated with chemoradiotherapy as the first-line treatment. However, diagnosis of a non-UIP pattern could predict lower risk of AE of ILD and longer OS durations.

A Deep Learning Approach For the Detection and Classification of Interstitial Lung Diseases Using Convolutional Neural Network

International Journal of Computer Sciences and Engineering (A UGC Approved and indexed with DOI, ICI and Approved, DPI Digital Library) is one of the leading and growing open access, peer-reviewed, monthly, and scientific research journal for scientists, engineers, research scholars, and academicians, which gains a foothold in Asia and opens to the world, aims to publish original, theoretical and practical advances in Computer Science,Information Technology, Engineering (Software, Mechanical, Civil, Electronics & Electrical), and all interdisciplinary streams of Computing Sciences. It intends to disseminate original, scientific, theoretical or applied research in the field of Computer Sciences and allied fields. It provides a platform for publishing results and research with a strong empirical component. It aims to bridge the significant gap between research and practice by promoting the publication of original, novel, industry-relevant research.