Classification of Interstitial Lung Abnormality Patterns with an Ensemble of Deep Convolutional Neural Networks

Automatic Quantification of Abnormal Lung Parenchymal Attenuation on Chest Computed Tomography Images Using Densitometry and Texture-based Analysis

PURPOSE

To compare texture-based analysis using convolutional neural networks (CNNs) against lung densitometry in detecting chest computed tomography (CT) image abnormalities.

MATERIAL AND METHODS

A U-NET was used for lung segmentation, and an ensemble of 7 CNN architectures was trained for the classification of low-attenuation areas (LAAs; emphysema, cysts), normal-attenuation areas (NAAs; normal parenchyma), and high-attenuation areas (HAAs; ground-glass opacities, crazy paving/linear opacity, consolidation). Lung densitometry also computes (LAAs, ≤-950 HU), NAAs (-949 to -700 HU), and HAAs (-699 to -250 HU). CNN-based and densitometry-based severity indices (CNN and Dens, respectively) were calculated as (LAA+HAA)/(LAA+NAA+HAA) in 812 CT scans from 176 normal subjects, 343 patients with emphysema, and 293 patients with interstitial lung disease (ILD). The correlation between CNN-derived and densitometry-derived indices was analyzed, alongside a comparison of severity indices among patient subgroups with emphysema and ILD, using the Spearman correlation and ANOVA with Bonferroni correction.

RESULTS

CNN-derived and densitometry-derived severity indices (SIs) showed a strong correlation (ρ=0.90) and increased with disease severity. CNN-SIs differed from densitometry SIs, being lower for emphysema and higher for moderate to severe ILD cases. CNN estimations for normal attenuation areas were higher than those from densitometry across all groups, indicating a potential for more accurate characterization of lung abnormalities.

CONCLUSIONS

CNN outputs align closely with densitometry in assessing lung abnormalities on CT scans, offering improved estimates of normal areas and better distinguishing similar abnormalities. However, this requires higher computing power.

Automated Interstitial Lung Abnormality Probability Prediction at CT: A Stepwise Machine Learning Approach in the Boston Lung Cancer Study

Background It is increasingly recognized that interstitial lung abnormalities (ILAs) detected at CT have potential clinical implications, but automated identification of ILAs has not yet been fully established. Purpose To develop and test automated ILA probability prediction models using machine learning techniques on CT images. Materials and Methods This secondary analysis of a retrospective study included CT scans from patients in the Boston Lung Cancer Study collected between February 2004 and June 2017. Visual assessment of ILAs by two radiologists and a pulmonologist served as the ground truth. Automated ILA probability prediction models were developed that used a stepwise approach involving section inference and case inference models. The section inference model produced an ILA probability for each CT section, and the case inference model integrated these probabilities to generate the case-level ILA probability. For indeterminate sections and cases, both two- and three-label methods were evaluated. For the case inference model, we tested three machine learning classifiers (support vector machine [SVM], random forest [RF], and convolutional neural network [CNN]). Receiver operating characteristic analysis was performed to calculate the area under the receiver operating characteristic curve (AUC). Results A total of 1382 CT scans (mean patient age, 67 years ± 11 [SD]; 759 women) were included. Of the 1382 CT scans, 104 (8%) were assessed as having ILA, 492 (36%) as indeterminate for ILA, and 786 (57%) as without ILA according to ground-truth labeling. The cohort was divided into a training set (n = 96; ILA, n = 48), a validation set (n = 24; ILA, n = 12), and a test set (n = 1262; ILA, n = 44). Among the models evaluated (two- and three-label section inference models; two- and three-label SVM, RF, and CNN case inference models), the model using the three-label method in the section inference model and the two-label method and RF in the case inference model achieved the highest AUC, at 0.87. Conclusion The model demonstrated substantial performance in estimating ILA probability, indicating its potential utility in clinical settings. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Zagurovskaya in this issue.

Weakly-supervised learning-based pathology detection and localization in 3D chest CT scans

BACKGROUND

Recent advancements in anomaly detection have paved the way for novel radiological reading assistance tools that support the identification of findings, aimed at saving time. The clinical adoption of such applications requires a low rate of false positives while maintaining high sensitivity.

PURPOSE

In light of recent interest and development in multi pathology identification, we present a novel method, based on a recent contrastive self-supervised approach, for multiple chest-related abnormality identification including low lung density area ("LLDA"), consolidation ("CONS"), nodules ("NOD") and interstitial pattern ("IP"). Our approach alerts radiologists about abnormal regions within a computed tomography (CT) scan by providing 3D localization.

METHODS

We introduce a new method for the classification and localization of multiple chest pathologies in 3D Chest CT scans. Our goal is to distinguish four common chest-related abnormalities: "LLDA", "CONS", "NOD", "IP" and "NORMAL". This method is based on a 3D patch-based classifier with a Resnet backbone encoder pretrained leveraging recent contrastive self supervised approach and a fine-tuned classification head. We leverage the SimCLR contrastive framework for pretraining on an unannotated dataset of randomly selected patches and we then fine-tune it on a labeled dataset. During inference, this classifier generates probability maps for each abnormality across the CT volume, which are aggregated to produce a multi-label patient-level prediction. We compare different training strategies, including random initialization, ImageNet weight initialization, frozen SimCLR pretrained weights and fine-tuned SimCLR pretrained weights. Each training strategy is evaluated on a validation set for hyperparameter selection and tested on a test set. Additionally, we explore the fine-tuned SimCLR pretrained classifier for 3D pathology localization and conduct qualitative evaluation.

RESULTS

Validated on 111 chest scans for hyperparameter selection and subsequently tested on 251 chest scans with multi-abnormalities, our method achieves an AUROC of 0.931 (95% confidence interval [CI]: [0.9034, 0.9557], p $ p$ -value < 0.001) and 0.963 (95% CI: [0.952, 0.976], p $ p$ -value < 0.001) in the multi-label and binary (i.e., normal versus abnormal) settings, respectively. Notably, our method surpasses the area under the receiver operating characteristic (AUROC) threshold of 0.9 for two abnormalities: IP (0.974) and LLDA (0.952), while achieving values of 0.853 and 0.791 for NOD and CONS, respectively. Furthermore, our results highlight the superiority of incorporating contrastive pretraining within the patch classifier, outperforming Imagenet pretraining weights and non-pretrained counterparts with uninitialized weights (F1 score = 0.943, 0.792, and 0.677 respectively). Qualitatively, the method achieved a satisfactory 88.8% completeness rate in localization and maintained an 88.3% accuracy rate against false positives.

CONCLUSIONS

The proposed method integrates self-supervised learning algorithms for pretraining, utilizes a patch-based approach for 3D pathology localization and develops an aggregation method for multi-label prediction at patient-level. It shows promise in efficiently detecting and localizing multiple anomalies within a single scan.

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.

Interstitial Lung Abnormalities: Current Understanding

Interstitial lung abnormalities (ILAs) are incidental findings on computed tomography scans, characterized by nondependent abnormalities affecting more than 5% of any lung zone. They are associated with factors such as age, smoking, genetic variants, worsened clinical outcomes, and increased mortality. Risk stratification based on clinical and radiological features of ILAs is crucial in clinical practice, particularly for identifying cases at high risk of progression to pulmonary fibrosis. Traction bronchiectasis/bronchiolectasis index has emerged as a promising imaging biomarker for prognostic risk stratification in ILAs. These findings suggest a spectrum of fibrosing interstitial lung diseases, encompassing from ILAs to pulmonary fibrosis.

Utility of peripheral protein biomarkers for the prediction of incident interstitial features: a multicentre retrospective cohort study

INTRODUCTION/RATIONALE

Protein biomarkers may help enable the prediction of incident interstitial features on chest CT.

METHODS

We identified which protein biomarkers in a cohort of smokers (COPDGene) differed between those with and without objectively measured interstitial features at baseline using a univariate screen (t-test false discovery rate, FDR p<0.001), and which of those were associated with interstitial features longitudinally (multivariable mixed effects model FDR p<0.05). To predict incident interstitial features, we trained four random forest classifiers in a two-thirds random subset of COPDGene: (1) imaging and demographic information, (2) univariate screen biomarkers, (3) multivariable confirmation biomarkers and (4) multivariable confirmation biomarkers available in a separate testing cohort (Pittsburgh Lung Screening Study (PLuSS)). We evaluated classifier performance in the remaining one-third of COPDGene, and, for the final model, also in PLuSS.

RESULTS

In COPDGene, 1305 biomarkers were available and 20 differed between those with and without interstitial features at baseline. Of these, 11 were associated with feature progression over a mean of 5.5 years of follow-up, and of these 4 were available in PLuSS, (angiopoietin-2, matrix metalloproteinase 7, macrophage inflammatory protein 1 alpha) over a mean of 8.8 years of follow-up. The area under the curve (AUC) of classifiers using demographics and imaging features in COPDGene and PLuSS were 0.69 and 0.59, respectively. In COPDGene, the AUC of the univariate screen classifier was 0.78 and of the multivariable confirmation classifier was 0.76. The AUC of the final classifier in COPDGene was 0.75 and in PLuSS was 0.76. The outcome for all of the models was the development of incident interstitial features.

CONCLUSIONS

Multiple novel and previously identified proteomic biomarkers are associated with interstitial features on chest CT and may enable the prediction of incident interstitial diseases such as idiopathic pulmonary fibrosis.

The influence of image reconstruction methods on the diagnosis of pulmonary emphysema with convolutional neural network

This study investigated the influence of iterative reconstruction (IR) methods on computed tomography (CT) images when training convolutional neural network (CNN) models to diagnose pulmonary emphysema. To evaluate the influence of the IR algorithm on CNN, the present study comprised two steps: the comparison of noise reduction by IR algorithms using phantom examinations and the change in performance of CNN with IR algorithms using patient data. We retrospectively analyzed 97 patients. Raw CT data were reconstructed using the filtered back-projection (FBP) and adaptive statistical iterative reconstruction V (ASIR-V) algorithms with blending levels of 30%, 50%, and 70%. The models were trained using reconstructed CT images and were named the FBP, ASIR-V30, ASIR-V50, and ASIR-V70 models. The mean and the standard deviation of the CT values were 11.3 ± 21.2 at FBP, 11.0 ± 17.3 at ASIR-V30, 11.0 ± 14.4 at ASIR-V50, and 11.0 ± 11.8 at ASIR-V70. For all the evaluation metrics, the best values were obtained with the FBP model applied to the ASIR-V70 test images. The worst values were obtained with the ASIR-V70 model applied to the FBP test images. The model trained with FBP images exhibited significantly better performance than the models trained using IR images. The reduction in image noise with the IR algorithm on the test images contributed to improving the accuracy of the classification of emphysema subtypes using CNN.

Recent Advancements in Computed Tomography Assessment of Fibrotic Interstitial Lung Diseases

Interstitial lung disease (ILD) is a heterogeneous group of disorders with complex and varied imaging manifestations and prognosis. High-resolution computed tomography (HRCT) is the current standard-of-care imaging tool for ILD assessment. However, visual evaluation of HRCT is limited by interobserver variation and poor sensitivity for subtle changes. Such challenges have led to tremendous recent research interest in objective and reproducible methods to examine ILDs. Computer-aided CT analysis to include texture analysis and machine learning methods have recently been shown to be viable supplements to traditional visual assessment through improved characterization and quantification of ILDs. These quantitative tools have not only been shown to correlate well with pulmonary function tests and patient outcomes but are also useful in disease diagnosis, surveillance and management. In this review, we provide an overview of recent computer-aided tools in diagnosis, prognosis, and longitudinal evaluation of fibrotic ILDs, while outlining some of the pitfalls and challenges that have precluded further advancement of these tools as well as potential solutions and further endeavors.

Deep Learning Assessment of Progression of Emphysema and Fibrotic Interstitial Lung Abnormality

Rationale: Although studies have evaluated emphysema and fibrotic interstitial lung abnormality individually, less is known about their combined progression. Objectives: To define clinically meaningful progression of fibrotic interstitial lung abnormality in smokers without interstitial lung disease and evaluate the effects of fibrosis and emphysema progression on mortality. Methods: Emphysema and pulmonary fibrosis were assessed on the basis of baseline and 5-year follow-up computed tomography scans of 4,450 smokers in the COPDGene Study using deep learning algorithms. Emphysema was classified as absent, trace, mild, moderate, confluent, or advanced destructive. Fibrosis was expressed as a percentage of lung volume. Emphysema progression was defined as an increase by at least one grade. A hybrid distribution and anchor-based method was used to determine the minimal clinically important difference in fibrosis. The relationship between progression and mortality was evaluated using multivariable shared frailty models using an age timescale. Measurements and Main Results: The minimal clinically important difference for fibrosis was 0.58%. On the basis of this threshold, 2,822 (63%) had progression of neither emphysema nor fibrosis, 841 (19%) had emphysema progression alone, 512 (12%) had fibrosis progression alone, and 275 (6.2%) had progression of both. Compared with nonprogressors, hazard ratios for mortality were 1.42 (95% confidence interval, 1.11-1.82) in emphysema progressors, 1.49 (1.14-1.94) in fibrosis progressors, and 2.18 (1.58-3.02) in those with progression of both emphysema and fibrosis. Conclusions: In smokers without known interstitial lung disease, small changes in fibrosis may be clinically significant, and combined progression of emphysema and fibrosis is associated with increased mortality.

Computer-Aided Pulmonary Fibrosis Detection Leveraging an Advanced Artificial Intelligence Triage and Notification Software

Background

Improvement in recognition and referral of pulmonary fibrosis (PF) is vital to improving patient outcomes within interstitial lung disease. We determined the performance metrics and processing time of an artificial intelligence triage and notification software, ScreenDx-LungFibrosis™, developed to improve detection of PF.

Methods

ScreenDx-LungFibrosis™ was applied to chest computed tomography (CT) scans from multisource data. Device output (+/- PF) was compared to clinical diagnosis (+/- PF), and diagnostic performance was evaluated. Primary endpoints included device sensitivity and specificity > 80% and processing time < 4.5 min.

Results

Of 3,018 patients included, PF was present in 22.9%. ScreenDx-LungFibrosis™ detected PF with a sensitivity and specificity of 91.3% (95% confidence interval (CI): 89.0-93.3%) and 95.1% (95% CI: 94.2-96.0%), respectively. Mean processing time was 27.6 s (95% CI: 26.0 - 29.1 s).

Conclusions

ScreenDx-LungFibrosis™ accurately and reliably identified PF with a rapid per-case processing time, underscoring its potential for transformative improvement in PF outcomes when routinely applied to chest CTs.

Lung Imaging in COPD Part 2: Emerging Concepts

The diagnosis, prognostication, and differentiation of phenotypes of COPD can be facilitated by CT scan imaging of the chest. CT scan imaging of the chest is a prerequisite for lung volume reduction surgery and lung transplantation. Quantitative analysis can be used to evaluate extent of disease progression. Evolving imaging techniques include micro-CT scan, ultra-high-resolution and photon-counting CT scan imaging, and MRI. Potential advantages of these newer techniques include improved resolution, prediction of reversibility, and obviation of radiation exposure. This article discusses important emerging techniques in imaging patients with COPD. The clinical usefulness of these emerging techniques as they stand today are tabulated for the benefit of the practicing pulmonologist.

Semi-Supervised Segmentation of Interstitial Lung Disease Patterns from CT Images via Self-Training with Selective Re-Training

Accurate segmentation of interstitial lung disease (ILD) patterns from computed tomography (CT) images is an essential prerequisite to treatment and follow-up. However, it is highly time-consuming for radiologists to pixel-by-pixel segment ILD patterns from CT scans with hundreds of slices. Consequently, it is hard to obtain large amounts of well-annotated data, which poses a huge challenge for data-driven deep learning-based methods. To alleviate this problem, we propose an end-to-end semi-supervised learning framework for the segmentation of ILD patterns (ESSegILD) from CT images via self-training with selective re-training. The proposed ESSegILD model is trained using a large CT dataset with slice-wise sparse annotations, i.e., only labeling a few slices in each CT volume with ILD patterns. Specifically, we adopt a popular semi-supervised framework, i.e., Mean-Teacher, that consists of a teacher model and a student model and uses consistency regularization to encourage consistent outputs from the two models under different perturbations. Furthermore, we propose introducing the latest self-training technique with a selective re-training strategy to select reliable pseudo-labels generated by the teacher model, which are used to expand training samples to promote the student model during iterative training. By leveraging consistency regularization and self-training with selective re-training, our proposed ESSegILD can effectively utilize unlabeled data from a partially annotated dataset to progressively improve the segmentation performance. Experiments are conducted on a dataset of 67 pneumonia patients with incomplete annotations containing over 11,000 CT images with eight different lung patterns of ILDs, with the results indicating that our proposed method is superior to the state-of-the-art methods.

Beyond Visual Interpretation: Quantitative Analysis and Artificial Intelligence in Interstitial Lung Disease Diagnosis "Expanding Horizons in Radiology"

Diffuse lung disorders (DLDs) and interstitial lung diseases (ILDs) are pathological conditions affecting the lung parenchyma and interstitial network. There are approximately 200 different entities within this category. Radiologists play an increasingly important role in diagnosing and monitoring ILDs, as they can provide non-invasive, rapid, and repeatable assessments using high-resolution computed tomography (HRCT). HRCT offers a detailed view of the lung parenchyma, resembling a low-magnification anatomical preparation from a histological perspective. The intrinsic contrast provided by air in HRCT enables the identification of even the subtlest morphological changes in the lung tissue. By interpreting the findings observed on HRCT, radiologists can make a differential diagnosis and provide a pattern diagnosis in collaboration with the clinical and functional data. The use of quantitative software and artificial intelligence (AI) further enhances the analysis of ILDs, providing an objective and comprehensive evaluation. The integration of "meta-data" such as demographics, laboratory, genomic, metabolomic, and proteomic data through AI could lead to a more comprehensive clinical and instrumental profiling beyond the human eye's capabilities.

Recent Advances of Artificial Intelligence Applications in Interstitial Lung Diseases

Interstitial lung diseases (ILDs) comprise a rather heterogeneous group of diseases varying in pathophysiology, presentation, epidemiology, diagnosis, treatment and prognosis. Even though they have been recognized for several years, there are still areas of research debate. In the majority of ILDs, imaging modalities and especially high-resolution Computed Tomography (CT) scans have been the cornerstone in patient diagnostic approach and follow-up. The intricate nature of ILDs and the accompanying data have led to an increasing adoption of artificial intelligence (AI) techniques, primarily on imaging data but also in genetic data, spirometry and lung diffusion, among others. In this literature review, we describe the most prominent applications of AI in ILDs presented approximately within the last five years. We roughly stratify these studies in three categories, namely: (i) screening, (ii) diagnosis and classification, (iii) prognosis.

Interstitial lung disease diagnosis and prognosis using an AI system integrating longitudinal data

For accurate diagnosis of interstitial lung disease (ILD), a consensus of radiologic, pathological, and clinical findings is vital. Management of ILD also requires thorough follow-up with computed tomography (CT) studies and lung function tests to assess disease progression, severity, and response to treatment. However, accurate classification of ILD subtypes can be challenging, especially for those not accustomed to reading chest CTs regularly. Dynamic models to predict patient survival rates based on longitudinal data are challenging to create due to disease complexity, variation, and irregular visit intervals. Here, we utilize RadImageNet pretrained models to diagnose five types of ILD with multimodal data and a transformer model to determine a patient's 3-year survival rate. When clinical history and associated CT scans are available, the proposed deep learning system can help clinicians diagnose and classify ILD patients and, importantly, dynamically predict disease progression and prognosis.

QUANTITATIVE BIOMARKERS REPRODUCIBILITY USING GENERATIVE ADVERSARIAL APPROACHES IN REDUCED TO CONVENTIONAL DOSE CT

In recent years, several techniques for image-to-image translation by means of generative adversarial neural networks (GAN) have been proposed to learn mapping characteristics between a source and a target domain. In particular, in the medical imaging field conditional GAN frameworks with paired samples (cGAN) and unconditional cycle-consistent GANs with unpaired data (CycleGAN) have been demonstrated as a powerful scheme to model non-linear mappings that produce realistic target images from different modality sources. When proposing the usage and adaptation of these frameworks for medical image synthesis, quantitative and qualitative validation are usually performed by assessing the similarity between synthetic and target images in terms of metrics such as mean absolute error (MAE) or structural similarity (SSIM) index. However, an evaluation of clinically relevant markers showing that diagnostic information is not overlooked in the translation process is often missing. In this work, we aim at demonstrating the importance of validating medical image-to-image translation techniques by assessing their effect on the measurement of clinically relevant metrics and biomarkers. We implemented both a conditional and an unconditional approach to synthesize conventional dose chest CT scans from reduced dose CT and show that while both visually and in terms of traditional metrics the network appears to successfully minimize perceptual discrepancies, these methods are not reliable to systematically reproduce quantitative measurements of various chest biomarkers.

AI co-pilot: content-based image retrieval for the reading of rare diseases in chest CT

The aim of the study was to evaluate the impact of the newly developed Similar patient search (SPS) Web Service, which supports reading complex lung diseases in computed tomography (CT), on the diagnostic accuracy of residents. SPS is an image-based search engine for pre-diagnosed cases along with related clinical reference content ( https://eref.thieme.de ). The reference database was constructed using 13,658 annotated regions of interest (ROIs) from 621 patients, comprising 69 lung diseases. For validation, 50 CT scans were evaluated by five radiology residents without SPS, and three months later with SPS. The residents could give a maximum of three diagnoses per case. A maximum of 3 points was achieved if the correct diagnosis without any additional diagnoses was provided. The residents achieved an average score of 17.6 ± 5.0 points without SPS. By using SPS, the residents increased their score by 81.8% to 32.0 ± 9.5 points. The improvement of the score per case was highly significant (p = 0.0001). The residents required an average of 205.9 ± 350.6 s per case (21.9% increase) when SPS was used. However, in the second half of the cases, after the residents became more familiar with SPS, this increase dropped to 7%. Residents' average score in reading complex chest CT scans improved by 81.8% when the AI-driven SPS with integrated clinical reference content was used. The increase in time per case due to the use of the SPS was minimal.

Artificial intelligence for differentiating COVID-19 from other viral pneumonias on CT: comparative analysis of different models based on quantitative and radiomic approaches

BACKGROUND

To develop a pipeline for automatic extraction of quantitative metrics and radiomic features from lung computed tomography (CT) and develop artificial intelligence (AI) models supporting differential diagnosis between coronavirus disease 2019 (COVID-19) and other viral pneumonia (non-COVID-19).

METHODS

Chest CT of 1,031 patients (811 for model building; 220 as independent validation set (IVS) with positive swab for severe acute respiratory syndrome coronavirus-2 (647 COVID-19) or other respiratory viruses (384 non-COVID-19) were segmented automatically. A Gaussian model, based on the HU histogram distribution describing well-aerated and ill portions, was optimised to calculate quantitative metrics (QM, n = 20) in both lungs (2L) and four geometrical subdivisions (GS) (upper front, lower front, upper dorsal, lower dorsal; n = 80). Radiomic features (RF) of first (RF1, n = 18) and second (RF2, n = 120) order were extracted from 2L using PyRadiomics tool. Extracted metrics were used to develop four multilayer-perceptron classifiers, built with different combinations of QM and RF: Model1 (RF1-2L); Model2 (QM-2L, QM-GS); Model3 (RF1-2L, RF2-2L); Model4 (RF1-2L, QM-2L, GS-2L, RF2-2L).

RESULTS

The classifiers showed accuracy from 0.71 to 0.80 and area under the receiving operating characteristic curve (AUC) from 0.77 to 0.87 in differentiating COVID-19 versus non-COVID-19 pneumonia. Best results were associated with Model3 (AUC 0.867 ± 0.008) and Model4 (AUC 0.870 ± 0.011. For the IVS, the AUC values were 0.834 ± 0.008 for Model3 and 0.828 ± 0.011 for Model4.

CONCLUSIONS

Four AI-based models for classifying patients as COVID-19 or non-COVID-19 viral pneumonia showed good diagnostic performances that could support clinical decisions.

Interstitial lung abnormalities in a large clinical lung cancer screening cohort: association with mortality and ILD diagnosis

BACKGROUND

Interstitial lung abnormalities (ILA) are CT findings suggestive of interstitial lung disease in individuals without a prior diagnosis or suspicion of ILD. Previous studies have demonstrated that ILA are associated with clinically significant outcomes including mortality. The aim of this study was to determine the prevalence of ILA in a large CT lung cancer screening program and the association with clinically significant outcomes including mortality, hospitalizations, cancer and ILD diagnosis.

METHODS

This was a retrospective study of individuals enrolled in a CT lung cancer screening program from 2012 to 2014. Baseline and longitudinal CT scans were scored for ILA per Fleischner Society guidelines. The primary analyses examined the association between baseline ILA and mortality, all-cause hospitalization, and incidence of lung cancer. Kaplan-Meier plots were generated to visualize the associations between ILA and lung cancer and all-cause mortality. Cox regression proportional hazards models were used to test for this association in both univariate and multivariable models.

RESULTS

1699 subjects met inclusion criteria. 41 (2.4%) had ILA and 101 (5.9%) had indeterminate ILA on baseline CTs. ILD was diagnosed in 10 (24.4%) of 41 with ILA on baseline CT with a mean time from baseline CT to diagnosis of 4.47 ± 2.72 years. On multivariable modeling, the presence of ILA remained a significant predictor of death, HR 3.87 (2.07, 7.21; p < 0.001) when adjusted for age, sex, BMI, pack years and active smoking, but not of lung cancer and all-cause hospital admission. Approximately 50% with baseline ILA had progression on the longitudinal scan.

CONCLUSIONS

ILA identified on baseline lung cancer screening exams are associated with all-cause mortality. In addition, a significant proportion of patients with ILA are subsequently diagnosed with ILD and have CT progression on longitudinal scans.

TRIAL REGISTRATION NUMBER

ClinicalTrials.gov; No.: NCT04503044.

Deep learning-based lesion subtyping and prediction of clinical outcomes in COVID-19 pneumonia using chest CT

The main objective of this work is to develop and evaluate an artificial intelligence system based on deep learning capable of automatically identifying, quantifying, and characterizing COVID-19 pneumonia patterns in order to assess disease severity and predict clinical outcomes, and to compare the prediction performance with respect to human reader severity assessment and whole lung radiomics. We propose a deep learning based scheme to automatically segment the different lesion subtypes in nonenhanced CT scans. The automatic lesion quantification was used to predict clinical outcomes. The proposed technique has been independently tested in a multicentric cohort of 103 patients, retrospectively collected between March and July of 2020. Segmentation of lesion subtypes was evaluated using both overlapping (Dice) and distance-based (Hausdorff and average surface) metrics, while the proposed system to predict clinically relevant outcomes was assessed using the area under the curve (AUC). Additionally, other metrics including sensitivity, specificity, positive predictive value and negative predictive value were estimated. 95% confidence intervals were properly calculated. The agreement between the automatic estimate of parenchymal damage (%) and the radiologists' severity scoring was strong, with a Spearman correlation coefficient (R) of 0.83. The automatic quantification of lesion subtypes was able to predict patient mortality, admission to the Intensive Care Units (ICU) and need for mechanical ventilation with an AUC of 0.87, 0.73 and 0.68 respectively. The proposed artificial intelligence system enabled a better prediction of those clinically relevant outcomes when compared to the radiologists' interpretation and to whole lung radiomics. In conclusion, deep learning lesion subtyping in COVID-19 pneumonia from noncontrast chest CT enables quantitative assessment of disease severity and better prediction of clinical outcomes with respect to whole lung radiomics or radiologists' severity score.

Idiopathic Pulmonary Fibrosis (an Update) and Progressive Pulmonary Fibrosis in Adults: An Official ATS/ERS/JRS/ALAT Clinical Practice Guideline

Background: This American Thoracic Society, European Respiratory Society, Japanese Respiratory Society, and Asociación Latinoamericana de Tórax guideline updates prior idiopathic pulmonary fibrosis (IPF) guidelines and addresses the progression of pulmonary fibrosis in patients with interstitial lung diseases (ILDs) other than IPF. Methods: A committee was composed of multidisciplinary experts in ILD, methodologists, and patient representatives. 1) Update of IPF: Radiological and histopathological criteria for IPF were updated by consensus. Questions about transbronchial lung cryobiopsy, genomic classifier testing, antacid medication, and antireflux surgery were informed by systematic reviews and answered with evidence-based recommendations using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach. 2) Progressive pulmonary fibrosis (PPF): PPF was defined, and then radiological and physiological criteria for PPF were determined by consensus. Questions about pirfenidone and nintedanib were informed by systematic reviews and answered with evidence-based recommendations using the GRADE approach. Results:1) Update of IPF: A conditional recommendation was made to regard transbronchial lung cryobiopsy as an acceptable alternative to surgical lung biopsy in centers with appropriate expertise. No recommendation was made for or against genomic classifier testing. Conditional recommendations were made against antacid medication and antireflux surgery for the treatment of IPF. 2) PPF: PPF was defined as at least two of three criteria (worsening symptoms, radiological progression, and physiological progression) occurring within the past year with no alternative explanation in a patient with an ILD other than IPF. A conditional recommendation was made for nintedanib, and additional research into pirfenidone was recommended. Conclusions: The conditional recommendations in this guideline are intended to provide the basis for rational, informed decisions by clinicians.

Assessing radiomics feature stability with simulated CT acquisitions

Medical imaging quantitative features had once disputable usefulness in clinical studies. Nowadays, advancements in analysis techniques, for instance through machine learning, have enabled quantitative features to be progressively useful in diagnosis and research. Tissue characterisation is improved via the “radiomics” features, whose extraction can be automated. Despite the advances, stability of quantitative features remains an important open problem. As features can be highly sensitive to variations of acquisition details, it is not trivial to quantify stability and efficiently select stable features. In this work, we develop and validate a Computed Tomography (CT) simulator environment based on the publicly available ASTRA toolbox (www.astra-toolbox.com). We show that the variability, stability and discriminative power of the radiomics features extracted from the virtual phantom images generated by the simulator are similar to those observed in a tandem phantom study. Additionally, we show that the variability is matched between a multi-center phantom study and simulated results. Consequently, we demonstrate that the simulator can be utilised to assess radiomics features’ stability and discriminative power.

Artificial Intelligence for Interstitial Lung Disease Analysis on Chest Computed Tomography: A Systematic Review

RATIONALE AND OBJECTIVES

High-resolution computed tomography (HRCT) is paramount in the assessment of interstitial lung disease (ILD). Yet, HRCT interpretation of ILDs may be hampered by inter- and intra-observer variability. Recently, artificial intelligence (AI) has revolutionized medical image analysis. This technology has the potential to advance patient care in ILD. We aimed to systematically evaluate the application of AI for the analysis of ILD in HRCT.

MATERIALS AND METHODS

We searched MEDLINE/PubMed databases for original publications of deep learning for ILD analysis on chest CT. The search included studies published up to March 1, 2021. The risk of bias evaluation included tailored Quality Assessment of Diagnostic Accuracy Studies and the modified Joanna Briggs Institute Critical Appraisal checklist.

RESULTS

Data was extracted from 19 retrospective studies. Deep learning techniques included detection, segmentation, and classification of ILD on HRCT. Most studies focused on the classification of ILD into different morphological patterns. Accuracies of 78%-91% were achieved. Two studies demonstrated near-expert performance for the diagnosis of idiopathic pulmonary fibrosis (IPF). The Quality Assessment of Diagnostic Accuracy Studies tool identified a high risk of bias in 15/19 (78.9%) of the studies.

CONCLUSION

AI has the potential to contribute to the radiologic diagnosis and classification of ILD. However, the accuracy performance is still not satisfactory, and research is limited by a small number of retrospective studies. Hence, the existing published data may not be sufficiently reliable. Only well-designed prospective controlled studies can accurately assess the value of existing AI tools for ILD evaluation.

Decision fusion in healthcare and medicine: a narrative review

OBJECTIVE

To provide an overview of the decision fusion (DF) technique and describe the applications of the technique in healthcare and medicine at prevention, diagnosis, treatment and administrative levels.

BACKGROUND

The rapid development of technology over the past 20 years has led to an explosion in data growth in various industries, like healthcare. Big data analysis within the healthcare systems is essential for arriving to a value-based decision over a period of time. Diversity and uncertainty in big data analytics have made it impossible to analyze data by using conventional data mining techniques and thus alternative solutions are required. DF is a form of data fusion techniques that could increase the accuracy of diagnosis and facilitate interpretation, summarization and sharing of information.

METHODS

We conducted a review of articles published between January 1980 and December 2020 from various databases such as Google Scholar, IEEE, PubMed, Science Direct, Scopus and web of science using the keywords decision fusion (DF), information fusion, healthcare, medicine and big data. A total of 141 articles were included in this narrative review.

CONCLUSIONS

Given the importance of big data analysis in reducing costs and improving the quality of healthcare; along with the potential role of DF in big data analysis, it is recommended to know the full potential of this technique including the advantages, challenges and applications of the technique before its use. Future studies should focus on describing the methodology and types of data used for its applications within the healthcare sector.

Special Convolutional Neural Network for Identification and Positioning of Interstitial Lung Disease Patterns in Computed Tomography Images

In this paper, automated detection of interstitial lung disease patterns in high resolution computed tomography images is achieved by developing a faster region-based convolutional network based detector with GoogLeNet as a backbone. GoogLeNet is simplified by removing few inception models and used as the backbone of the detector network. The proposed framework is developed to detect several interstitial lung disease patterns without doing lung field segmentation. The proposed method is able to detect the five most prevalent interstitial lung disease patterns: fibrosis, emphysema, consolidation, micronodules and ground-glass opacity, as well as normal. Five-fold cross-validation has been used to avoid bias and reduce over-fitting. The proposed framework performance is measured in terms of F-score on the publicly available MedGIFT database. It outperforms state-of-the-art techniques. The detection is performed at slice level and could be used for screening and differential diagnosis of interstitial lung disease patterns using high resolution computed tomography images.

Interstitial lung abnormalities - current knowledge and future directions

Efforts to grasp the significance of radiologic changes similar to interstitial lung disease (ILD) in undiagnosed individuals have intensified in the recent decade. The term interstitial lung abnormalities (ILA) is an emerging definition of such changes, defined by visual examination of computed tomography scans. Substantial insights have been made in the origins and clinical consequences of these changes, as well as automated measures of early lung fibrosis, which will likely lead to increased recognition of early fibrotic lung changes among clinicians and researchers alike.

Interstitial lung abnormalities have an estimated prevalence of 7–10% in elderly populations. They correlate with many ILD risk factors, both epidemiologic and genetic. Additionally, histopathological similarities with IPF exist in those with ILA. While no established blood biomarker of ILA exists, several have been suggested. Distinct imaging patterns indicating advanced fibrosis correlate with worse clinical outcomes. ILA are also linked with adverse clinical outcomes such as increased mortality and risk of lung cancer.

Progression of ILA has been noted in a significant portion of those with ILA and is associated with many of the same features as ILD, including advanced fibrosis. Those with ILA progression are at risk of accelerated FVC decline and increased mortality. Radiologic changes resembling ILD have also been attained by automated measures. Such measures associate with some, but not all the same factors as ILA.

ILA and similar radiologic changes are in many ways analogous to ILD and likely represent a precursor of ILD in some cases. While warranting an evaluation for ILD, they are associated with poor clinical outcomes beyond possible ILD development and thus are by themselves a significant finding. Among the present objectives of this field are the stratification of patients with regards to progression and the discovery of biomarkers with predictive value for clinical outcomes.

Fibrosis-Net: A Tailored Deep Convolutional Neural Network Design for Prediction of Pulmonary Fibrosis Progression From Chest CT Images

Pulmonary fibrosis is a devastating chronic lung disease that causes irreparable lung tissue scarring and damage, resulting in progressive loss in lung capacity and has no known cure. A critical step in the treatment and management of pulmonary fibrosis is the assessment of lung function decline, with computed tomography (CT) imaging being a particularly effective method for determining the extent of lung damage caused by pulmonary fibrosis. Motivated by this, we introduce Fibrosis-Net, a deep convolutional neural network design tailored for the prediction of pulmonary fibrosis progression from chest CT images. More specifically, machine-driven design exploration was leveraged to determine a strong architectural design for CT lung analysis, upon which we build a customized network design tailored for predicting forced vital capacity (FVC) based on a patient's CT scan, initial spirometry measurement, and clinical metadata. Finally, we leverage an explainability-driven performance validation strategy to study the decision-making behavior of Fibrosis-Net as to verify that predictions are based on relevant visual indicators in CT images. Experiments using a patient cohort from the OSIC Pulmonary Fibrosis Progression Challenge showed that the proposed Fibrosis-Net is able to achieve a significantly higher modified Laplace Log Likelihood score than the winning solutions on the challenge. Furthermore, explainability-driven performance validation demonstrated that the proposed Fibrosis-Net exhibits correct decision-making behavior by leveraging clinically-relevant visual indicators in CT images when making predictions on pulmonary fibrosis progress. Fibrosis-Net is able to achieve a significantly higher modified Laplace Log Likelihood score than the winning solutions on the OSIC Pulmonary Fibrosis Progression Challenge, and has been shown to exhibit correct decision-making behavior when making predictions. Fibrosis-Net is available to the general public in an open-source and open access manner as part of the OpenMedAI initiative. While Fibrosis-Net is not yet a production-ready clinical assessment solution, we hope that its release will encourage researchers, clinicians, and citizen data scientists alike to leverage and build upon it.

Interstitial lung abnormalities: What do we know and how do we manage?

INTRODUCTION

Interstitial lung abnormalities (ILAs), which refer to mild or subtle nongravity-dependent interstitial changes, may be neglected by some clinicians due to many reasons, such as lack of diagnostic criteria for ILAs and absence of available treatments and surveillance strategies. However, without intervention, some ILAs may progress to interstitial lung disease (ILD). This review summarizes our current knowledge of this condition and ways of diagnosing it together with current management. We hope that this will lead to better recognition of ILAs.

AREAS COVERED

We reviewed the literature on PubMed between 2008 and 2020 focusing on prevalence, etiology, symptoms, diagnostic biomarkers, clinical associations, and management of ILAs.

EXPERT OPINION

Timely diagnosis with close monitoring of ILAs and appropriate intervention should be recognized as the management approach to ILAs. Research into ILAs should continue to improve its management.

Pulmonary Emphysema Regional Distribution and Extent Assessed by Chest Computed Tomography Is Associated With Pulmonary Function Impairment in Patients With COPD

Objective: This study aimed to evaluate how emphysema extent and its regional distribution quantified by chest CT are associated with clinical and functional severity in patients with chronic obstructive pulmonary disease (COPD). Methods/Design: Patients with a post-bronchodilator forced expiratory volume in one second (FEV1)/forced vital capacity (FVC) < 0.70, without any other obstructive airway disease, who presented radiological evidence of emphysema on visual CT inspection were retrospectively enrolled. A Quantitative Lung Imaging (QUALI) system automatically quantified the volume of pulmonary emphysema and adjusted this volume to the measured (EmphCTLV) or predicted total lung volume (TLV) (EmphPLV) and assessed its regional distribution based on an artificial neural network (ANN) trained for this purpose. Additionally, the percentage of lung volume occupied by low-attenuation areas (LAA) was computed by dividing the total volume of regions with attenuation lower or equal to -950 Hounsfield units (HU) by the predicted [LAA (%PLV)] or measured CT lung volume [LAA (%CTLV)]. The LAA was then compared with the QUALI emphysema estimations. The association between emphysema extension and its regional distribution with pulmonary function impairment was then assessed. Results: In this study, 86 patients fulfilled the inclusion criteria. Both EmphCTLV and EmphPLV were significantly lower than the LAA indices independently of emphysema severity. CT-derived TLV significantly increased with emphysema severity (from 6,143 ± 1,295 up to 7,659 ± 1,264 ml from mild to very severe emphysema, p < 0.005) and thus, both EmphCTLV and LAA significantly underestimated emphysema extent when compared with those values adjusted to the predicted lung volume. All CT-derived emphysema indices presented moderate to strong correlations with residual volume (RV) (with correlations ranging from 0.61 to 0.66), total lung capacity (TLC) (from 0.51 to 0.59), and FEV1 (~0.6) and diffusing capacity for carbon monoxide DLCO (~0.6). The values of FEV1 and DLCO were significantly lower, and RV (p < 0.001) and TLC (p < 0.001) were significantly higher with the increasing emphysema extent and when emphysematous areas homogeneously affected the lungs. Conclusions: Emphysema volume must be referred to the predicted and not to the measured lung volume when assessing the CT-derived emphysema extension. Pulmonary function impairment was greater in patients with higher emphysema volumes and with a more homogeneous emphysema distribution. Further studies are still necessary to assess the significance of CTpLV in the clinical and research fields.

An explainable AI system for automated COVID-19 assessment and lesion categorization from CT-scans

COVID-19 infection caused by SARS-CoV-2 pathogen has been a catastrophic pandemic outbreak all over the world, with exponential increasing of confirmed cases and, unfortunately, deaths. In this work we propose an AI-powered pipeline, based on the deep-learning paradigm, for automated COVID-19 detection and lesion categorization from CT scans. We first propose a new segmentation module aimed at automatically identifying lung parenchyma and lobes. Next, we combine the segmentation network with classification networks for COVID-19 identification and lesion categorization. We compare the model's classification results with those obtained by three expert radiologists on a dataset of 166 CT scans. Results showed a sensitivity of 90.3% and a specificity of 93.5% for COVID-19 detection, at least on par with those yielded by the expert radiologists, and an average lesion categorization accuracy of about 84%. Moreover, a significant role is played by prior lung and lobe segmentation, that allowed us to enhance classification performance by over 6 percent points. The interpretation of the trained AI models reveals that the most significant areas for supporting the decision on COVID-19 identification are consistent with the lesions clinically associated to the virus, i.e., crazy paving, consolidation and ground glass. This means that the artificial models are able to discriminate a positive patient from a negative one (both controls and patients with interstitial pneumonia tested negative to COVID) by evaluating the presence of those lesions into CT scans. Finally, the AI models are integrated into a user-friendly GUI to support AI explainability for radiologists, which is publicly available at http://perceivelab.com/covid-ai. The whole AI system is unique since, to the best of our knowledge, it is the first AI-based software, publicly available, that attempts to explain to radiologists what information is used by AI methods for making decisions and that proactively involves them in the decision loop to further improve the COVID-19 understanding.

Classification of microcalcification clusters in digital breast tomosynthesis using ensemble convolutional neural network

BACKGROUND

The classification of benign and malignant microcalcification clusters (MCs) is an important task for computer-aided diagnosis (CAD) of digital breast tomosynthesis (DBT) images. Influenced by imaging method, DBT has the characteristic of anisotropic resolution, in which the resolution of intra-slice and inter-slice is quite different. In addition, the sharpness of MCs in different slices of DBT is quite different, among which the clearest slice is called focus slice. These characteristics limit the performance of CAD algorithms based on standard 3D convolution neural network (CNN).

METHODS

To make full use of the characteristics of the DBT, we proposed a new ensemble CNN, which consists of the 2D ResNet34 and the anisotropic 3D ResNet to extract the 2D focus slice features and 3D contextual features of MCs, respectively. Moreover, the anisotropic 3D convolution is used to build 3D ResNet to avoid the influence of DBT anisotropy.

RESULTS

The proposed method was evaluated on 495 MCs in DBT images of 275 patients, which are collected from our collaborative hospital. The area under the curve (AUC) of receiver operating characteristic (ROC) and accuracy of classifying benign and malignant MCs using decision-level ensemble strategy were 0.8837 and 82.00%, which were significantly higher than the experimental results of 2D ResNet34 (AUC: 0.8264, ACC: 76.00%) and anisotropic 3D ResNet (AUC: 0.8455, ACC: 76.00%). Compared with the results of 3D features classification in the radiomics, the AUC of the deep learning method with decision-level ensemble strategy was improved by 0.0435, and the F1 score was improved from 79.37 to 85.71%. More importantly, the sensitivity increased from 78.13 to 84.38%, and the specificity increased from 66.67 to 77.78%, which effectively reduced the false positives of diagnosis CONCLUSION: The results fully prove that the ensemble CNN can effectively integrate 2D features and 3D features, improve the classification performance of benign and malignant MCs in DBT, and reduce the false positives.

Multiclass wound image classification using an ensemble deep CNN-based classifier

Acute and chronic wounds are a challenge to healthcare systems around the world and affect many people's lives annually. Wound classification is a key step in wound diagnosis that would help clinicians to identify an optimal treatment procedure. Hence, having a high-performance classifier assists wound specialists to classify wound types with less financial and time costs. Different wound classification methods based on machine learning and deep learning have been proposed in the literature. In this study, we have developed an ensemble Deep Convolutional Neural Network-based classifier to categorize wound images into multiple classes including surgical, diabetic, and venous ulcers. The output classification scores of two classifiers (namely, patch-wise and image-wise) are fed into a Multilayer Perceptron to provide a superior classification performance. A 5-fold cross-validation approach is used to evaluate the proposed method. We obtained maximum and average classification accuracy values of 96.4% and 94.28% for binary and 91.9% and 87.7% for 3-class classification problems. The proposed classifier was compared with some common deep classifiers and showed significantly higher accuracy metrics. We also tested the proposed method on the Medetec wound image dataset, and the accuracy values of 91.2% and 82.9% were obtained for binary and 3-class classifications. The results show that our proposed method can be used effectively as a decision support system in classification of wound images or other related clinical applications.

Automatic classification of heterogeneous slit-illumination images using an ensemble of cost-sensitive convolutional neural networks

Background

Lens opacity seriously affects the visual development of infants. Slit-illumination images play an irreplaceable role in lens opacity detection; however, these images exhibited varied phenotypes with severe heterogeneity and complexity, particularly among pediatric cataracts. Therefore, it is urgently needed to explore an effective computer-aided method to automatically diagnose heterogeneous lens opacity and to provide appropriate treatment recommendations in a timely manner.

Methods

We integrated three different deep learning networks and a cost-sensitive method into an ensemble learning architecture, and then proposed an effective model called CCNN-Ensemble [ensemble of cost-sensitive convolutional neural networks (CNNs)] for automatic lens opacity detection. A total of 470 slit-illumination images of pediatric cataracts were used for training and comparison between the CCNN-Ensemble model and conventional methods. Finally, we used two external datasets (132 independent test images and 79 Internet-based images) to further evaluate the model's generalizability and effectiveness.

Results

Experimental results and comparative analyses demonstrated that the proposed method was superior to conventional approaches and provided clinically meaningful performance in terms of three grading indices of lens opacity: area (specificity and sensitivity; 92.00% and 92.31%), density (93.85% and 91.43%) and opacity location (95.25% and 89.29%). Furthermore, the comparable performance on the independent testing dataset and the internet-based images verified the effectiveness and generalizability of the model. Finally, we developed and implemented a website-based automatic diagnosis software for pediatric cataract grading diagnosis in ophthalmology clinics.

Conclusions

The CCNN-Ensemble method demonstrates higher specificity and sensitivity than conventional methods on multi-source datasets. This study provides a practical strategy for heterogeneous lens opacity diagnosis and has the potential to be applied to the analysis of other medical images.

[Radiological features of interstitial lung diseases]

In addition to pneumology and pathology, radiology is an essential discipline in the interdisciplinary diagnosis of interstitial lung diseases (ILDs). The gold standard for diagnosis of ILD is computed tomography. Diagnostic findings are based on specific radiological signs such as interlobular septal thickening and nodular changes. From these signs and their distribution within the lung, radiological patterns can be derived, e.g., usual interstitial pneumonia, nonspecific interstitial pneumonia, or organizing pneumonia. Various differential diagnoses result from the radiological pattern, which can then be further limited in an interdisciplinary manner with the clinic and pathology and, if necessary, trigger further diagnostics.The visual assessment of interstitial lung changes requires experience and training and is nevertheless error-prone with high inter- and intraobserver variabilities. Recently, therefore, computer-aided analysis of ILDs has been increasingly promoted. These computer programs analyze the density distribution of the lung parenchyma using parameters such as mean lung density, skewness, and kurtosis thus enabling the quantification and assessment of the course of disease. Furthermore, texture analysis and artificial intelligence are used to characterize parenchymal changes and differentiate between regions of ground glass, reticulation, and honeycombing. Modern dual-energy CT methods allow a combined, regional recording of both the morphology and the function and provide information about regional ventilation and perfusion.

Automated detection of COVID-19 using ensemble of transfer learning with deep convolutional neural network based on CT scans

PURPOSE

COVID-19 has infected millions of people worldwide. One of the most important hurdles in controlling the spread of this disease is the inefficiency and lack of medical tests. Computed tomography (CT) scans are promising in providing accurate and fast detection of COVID-19. However, determining COVID-19 requires highly trained radiologists and suffers from inter-observer variability. To remedy these limitations, this paper introduces an automatic methodology based on an ensemble of deep transfer learning for the detection of COVID-19.

METHODS

A total of 15 pre-trained convolutional neural networks (CNNs) architectures: EfficientNets(B0-B5), NasNetLarge, NasNetMobile, InceptionV3, ResNet-50, SeResnet 50, Xception, DenseNet121, ResNext50 and Inception_resnet_v2 are used and then fine-tuned on the target task. After that, we built an ensemble method based on majority voting of the best combination of deep transfer learning outputs to further improve the recognition performance. We have used a publicly available dataset of CT scans, which consists of 349 CT scans labeled as being positive for COVID-19 and 397 negative COVID-19 CT scans that are normal or contain other types of lung diseases.

RESULTS

The experimental results indicate that the majority voting of 5 deep transfer learning architecture with EfficientNetB0, EfficientNetB3, EfficientNetB5, Inception_resnet_v2, and Xception has the higher results than the individual transfer learning structure and among the other models based on precision (0.857), recall (0.854) and accuracy (0.85) metrics in diagnosing COVID-19 from CT scans.

CONCLUSION

Our study based on an ensemble deep transfer learning system with different pre-trained CNNs architectures can work well on a publicly available dataset of CT images for the diagnosis of COVID-19 based on CT scans.

A deep transfer learning model with classical data augmentation and CGAN to detect COVID-19 from chest CT radiography digital images

The Coronavirus disease 2019 (COVID-19) is the fastest transmittable virus caused by severe acute respiratory syndrome Coronavirus 2 (SARS-CoV-2). The detection of COVID-19 using artificial intelligence techniques and especially deep learning will help to detect this virus in early stages which will reflect in increasing the opportunities of fast recovery of patients worldwide. This will lead to release the pressure off the healthcare system around the world. In this research, classical data augmentation techniques along with Conditional Generative Adversarial Nets (CGAN) based on a deep transfer learning model for COVID-19 detection in chest CT scan images will be presented. The limited benchmark datasets for COVID-19 especially in chest CT images are the main motivation of this research. The main idea is to collect all the possible images for COVID-19 that exists until the very writing of this research and use the classical data augmentations along with CGAN to generate more images to help in the detection of the COVID-19. In this study, five different deep convolutional neural network-based models (AlexNet, VGGNet16, VGGNet19, GoogleNet, and ResNet50) have been selected for the investigation to detect the Coronavirus-infected patient using chest CT radiographs digital images. The classical data augmentations along with CGAN improve the performance of classification in all selected deep transfer models. The outcomes show that ResNet50 is the most appropriate deep learning model to detect the COVID-19 from limited chest CT dataset using the classical data augmentation with testing accuracy of 82.91%, sensitivity 77.66%, and specificity of 87.62%.

AI-driven quantification, staging and outcome prediction of COVID-19 pneumonia

Coronavirus disease 2019 (COVID-19) emerged in 2019 and disseminated around the world rapidly. Computed tomography (CT) imaging has been proven to be an important tool for screening, disease quantification and staging. The latter is of extreme importance for organizational anticipation (availability of intensive care unit beds, patient management planning) as well as to accelerate drug development through rapid, reproducible and quantified assessment of treatment response. Even if currently there are no specific guidelines for the staging of the patients, CT together with some clinical and biological biomarkers are used. In this study, we collected a multi-center cohort and we investigated the use of medical imaging and artificial intelligence for disease quantification, staging and outcome prediction. Our approach relies on automatic deep learning-based disease quantification using an ensemble of architectures, and a data-driven consensus for the staging and outcome prediction of the patients fusing imaging biomarkers with clinical and biological attributes. Highly promising results on multiple external/independent evaluation cohorts as well as comparisons with expert human readers demonstrate the potentials of our approach.

Machine-learning-based multiple abnormality prediction with large-scale chest computed tomography volumes

Machine learning models for radiology benefit from large-scale data sets with high quality labels for abnormalities. We curated and analyzed a chest computed tomography (CT) data set of 36,316 volumes from 19,993 unique patients. This is the largest multiply-annotated volumetric medical imaging data set reported. To annotate this data set, we developed a rule-based method for automatically extracting abnormality labels from free-text radiology reports with an average F-score of 0.976 (min 0.941, max 1.0). We also developed a model for multi-organ, multi-disease classification of chest CT volumes that uses a deep convolutional neural network (CNN). This model reached a classification performance of AUROC >0.90 for 18 abnormalities, with an average AUROC of 0.773 for all 83 abnormalities, demonstrating the feasibility of learning from unfiltered whole volume CT data. We show that training on more labels improves performance significantly: for a subset of 9 labels - nodule, opacity, atelectasis, pleural effusion, consolidation, mass, pericardial effusion, cardiomegaly, and pneumothorax - the model's average AUROC increased by 10% when the number of training labels was increased from 9 to all 83. All code for volume preprocessing, automated label extraction, and the volume abnormality prediction model is publicly available. The 36,316 CT volumes and labels will also be made publicly available pending institutional approval.

Interstitial lung abnormalities detected incidentally on CT: a Position Paper from the Fleischner Society

The term interstitial lung abnormalities refers to specific CT findings that are potentially compatible with interstitial lung disease in patients without clinical suspicion of the disease. Interstitial lung abnormalities are increasingly recognised as a common feature on CT of the lung in older individuals, occurring in 4-9% of smokers and 2-7% of non-smokers. Identification of interstitial lung abnormalities will increase with implementation of lung cancer screening, along with increased use of CT for other diagnostic purposes. These abnormalities are associated with radiological progression, increased mortality, and the risk of complications from medical interventions, such as chemotherapy and surgery. Management requires distinguishing interstitial lung abnormalities that represent clinically significant interstitial lung disease from those that are subclinical. In particular, it is important to identify the subpleural fibrotic subtype, which is more likely to progress and to be associated with mortality. This multidisciplinary Position Paper by the Fleischner Society addresses important issues regarding interstitial lung abnormalities, including standardisation of the definition and terminology; predisposing risk factors; clinical outcomes; options for initial evaluation, monitoring, and management; the role of quantitative evaluation; and future research needs.