BM-BronchoLC - A rich bronchoscopy dataset for anatomical landmarks and lung cancer lesion recognition.

Flexible bronchoscopy has revolutionized respiratory disease diagnosis. It offers direct visualization and detection of airway abnormalities, including lung cancer lesions. Accurate identification of airway lesions during flexible bronchoscopy plays an important role in the lung cancer diagnosis. The application of artificial intelligence (AI) aims to support physicians in recognizing anatomical landmarks and lung cancer lesions within bronchoscopic imagery. This work described the development of BM-BronchoLC, a rich bronchoscopy dataset encompassing 106 lung cancer and 102 non-lung cancer patients. The dataset incorporates detailed localization and categorical annotations for both anatomical landmarks and lesions, meticulously conducted by senior doctors at Bach Mai Hospital, Vietnam. To assess the dataset's quality, we evaluate two prevalent AI backbone models, namely UNet++ and ESFPNet, on the image segmentation and classification tasks with single-task and multi-task learning paradigms. We present BM-BronchoLC as a reference dataset in developing AI models to assist diagnostic accuracy for anatomical landmarks and lung cancer lesions in bronchoscopy data.

Generalizable disease detection using model ensemble on chest X-ray images.

In the realm of healthcare, the demand for swift and precise diagnostic tools has been steadily increasing. This study delves into a comprehensive performance analysis of three pre-trained convolutional neural network (CNN) architectures: ResNet50, DenseNet121, and Inception-ResNet-v2. To ensure the broad applicability of our approach, we curated a large-scale dataset comprising a diverse collection of chest X-ray images, that included both positive and negative cases of COVID-19. The models' performance was evaluated using separate datasets for internal validation (from the same source as the training images) and external validation (from different sources). Our examination uncovered a significant drop in network efficacy, registering a 10.66% reduction for ResNet50, a 36.33% decline for DenseNet121, and a 19.55% decrease for Inception-ResNet-v2 in terms of accuracy. Best results were obtained with DenseNet121 achieving the highest accuracy at 96.71% in internal validation and Inception-ResNet-v2 attaining 76.70% accuracy in external validation. Furthermore, we introduced a model ensemble approach aimed at improving network performance when making inferences on images from diverse sources beyond their training data. The proposed method uses uncertainty-based weighting by calculating the entropy in order to assign appropriate weights to the outputs of each network. Our results showcase the effectiveness of the ensemble method in enhancing accuracy up to 97.38% for internal validation and 81.18% for external validation, while maintaining a balanced ability to detect both positive and negative cases.

Portable Dynamic Chest Radiography: Literature Review and Potential Bedside Applications.

Dynamic digital radiography (DDR) is a high-resolution radiographic imaging technique using pulsed X-ray emission to acquire a multiframe cine-loop of the target anatomical area. The first DDR technology was orthostatic chest acquisitions, but new portable equipment that can be positioned at the patient's bedside was recently released, significantly expanding its potential applications, particularly in chest examination. It provides anatomical and functional information on the motion of different anatomical structures, such as the lungs, pleura, rib cage, and trachea. Native images can be further analyzed with dedicated post-processing software to extract quantitative parameters, including diaphragm motility, automatically projected lung area and area changing rate, a colorimetric map of the signal value change related to respiration and motility, and lung perfusion. The dynamic diagnostic information along with the significant advantages of this technique in terms of portability, versatility, and cost-effectiveness represents a potential game changer for radiological diagnosis and monitoring at the patient's bedside. DDR has several applications in daily clinical practice, and in this narrative review, we will focus on chest imaging, which is the main application explored to date in the literature. However, studies are still needed to understand deeply the clinical impact of this method.

Image quality in CT thorax: effect of altering reconstruction algorithm and tube load.

Non-linear properties of iterative reconstruction (IR) algorithms can alter image texture. We evaluated the effect of a model-based IR algorithm (advanced modelled iterative reconstruction; ADMIRE) and dose on computed tomography thorax image quality. Dual-source scanner data were acquired at 20, 45 and 65 reference mAs in 20 patients. Images reconstructed with filtered back projection (FBP) and ADMIRE Strengths 3-5 were assessed independently by six radiologists and analysed using an ordinal logistic regression model. For all image criteria studied, the effects of tube load 20 mAs and all ADMIRE strengths were significant (p < 0.001) when compared to reference categories 65 mAs and FBP. Increase in tube load from 45 to 65 mAs showed image quality improvement in three of six criteria. Replacing FBP with ADMIRE significantly improves perceived image quality for all criteria studied, potentially permitting a dose reduction of almost 70% without loss in image quality.

Lung-DT: An AI-Powered Digital Twin Framework for Thoracic Health Monitoring and Diagnosis.

The integration of artificial intelligence (AI) with Digital Twins (DTs) has emerged as a promising approach to revolutionize healthcare, particularly in terms of diagnosis and management of thoracic disorders. This study proposes a comprehensive framework, named Lung-DT, which leverages IoT sensors and AI algorithms to establish the digital representation of a patient's respiratory health. Using the YOLOv8 neural network, the Lung-DT system accurately classifies chest X-rays into five distinct categories of lung diseases, including "normal", "covid", "lung_opacity", "pneumonia", and "tuberculosis". The performance of the system was evaluated employing a chest X-ray dataset available in the literature, demonstrating average accuracy of 96.8%, precision of 92%, recall of 97%, and F1-score of 94%. The proposed Lung-DT framework offers several advantages over conventional diagnostic methods. Firstly, it enables real-time monitoring of lung health through continuous data acquisition from IoT sensors, facilitating early diagnosis and intervention. Secondly, the AI-powered classification module provides automated and objective assessments of chest X-rays, reducing dependence on subjective human interpretation. Thirdly, the twin digital representation of the patient's respiratory health allows for comprehensive analysis and correlation of multiple data streams, providing valuable insights as to personalized treatment plans. The integration of IoT sensors, AI algorithms, and DT technology within the Lung-DT system demonstrates a significant step towards improving thoracic healthcare. By enabling continuous monitoring, automated diagnosis, and comprehensive data analysis, the Lung-DT framework has enormous potential to enhance patient outcomes, reduce healthcare costs, and optimize resource allocation.

CT-FEM of the human thorax: Frequency response function and 3D harmonic analysis at resonance.

BACKGROUND AND OBJECTIVE: High-frequency chest wall compression (HFCC) therapy by airway clearance devices (ACDs) acts on the rheological properties of bronchial mucus to assist in clearing pulmonary secretions. Investigating low-frequency vibrations on the human thorax through numerical simulations is critical to ensure consistency and repeatability of studies by reducing extreme variability in body measurements across individuals. This study aims to present the numerical investigation of the harmonic acoustic excitation of ACDs on the human chest as a gentle and effective HFCC therapy. METHODS: Four software programs were sequentially used to visualize medical images, decrease the number of surfaces, generate and repair meshes, and conduct numerical analysis, respectively. The developed methodology supplied the validation of the effect of HFCC through computed tomography-based finite element analysis (CT-FEM) of a human thorax. To illustrate the vibroacoustic characteristics of the HFCC therapy device, a 146-decibel sound pressure level (dBSPL) was applied on the back-chest surface of the model. Frequency response function (FRF) across 5-100 Hz was analyzed to characterize the behaviour of the human thorax with the state-space model. RESULTS: We discovered that FRF pertaining to accelerance equals 0.138 m/s2N at the peak frequency of 28 Hz, which is consistent with two independent experimental airway clearance studies reported in the literature. The state-space model assessed two apparent resonance frequencies at 28 Hz and 41 Hz for the human thorax. The total displacement, kinetic energy density, and elastic strain energy density were furthermore quantified at 1 µm, 5.2 µJ/m3, and 140.7 µJ/m3, respectively, at the resonance frequency. In order to deepen our understanding of the impact on internal organs, the model underwent simulations in both the time domain and frequency domain for a comprehensive analysis. CONCLUSION: Overall, the present study enabled determining and validating FRF of the human thorax to roll out the inconsistencies, contributing to the health of individuals with investigating gentle but effective HFCC therapy conditions with ACDs. This innovative finding furthermore provides greater clarity and a tangible understanding of the subject by simulating the responses of CT-FEM of the human thorax and internal organs at resonance.

Domain adaptation via Wasserstein distance and discrepancy metric for chest X-ray image classification.

Deep learning technology can effectively assist physicians in diagnosing chest radiographs. Conventional domain adaptation methods suffer from inaccurate lesion region localization, large errors in feature extraction, and a large number of model parameters. To address these problems, we propose a novel domain-adaptive method WDDM to achieve abnormal identification of chest radiographic images by combining Wasserstein distance and difference measures. Specifically, our method uses BiFormer as a multi-scale feature extractor to extract deep feature representations of data samples, which focuses more on discriminant features than convolutional neural networks and Swin Transformer. In addition, based on the loss minimization of Wasserstein distance and contrast domain differences, the source domain samples closest to the target domain are selected to achieve similarity and dissimilarity across domains. Experimental results show that compared with the non-transfer method that directly uses the network trained in the source domain to classify the target domain, our method has an average AUC increase of 14.8% and above. In short, our method achieves higher classification accuracy and better generalization performance.

Label correlation guided discriminative label feature learning for multi-label chest image classification.

BACKGROUND AND OBJECTIVE: Multi-label Chest X-ray (CXR) images often contain rich label relationship information, which is beneficial to improve classification performance. However, because of the intricate relationships among labels, most existing works fail to effectively learn and make full use of the label correlations, resulting in limited classification performance. In this study, we propose a multi-label learning framework that learns and leverages the label correlations to improve multi-label CXR image classification. METHODS: In this paper, we capture the global label correlations through the self-attention mechanism. Meanwhile, to better utilize label correlations for guiding feature learning, we decompose the image-level features into label-level features. Furthermore, we enhance label-level feature learning in an end-to-end manner by a consistency constraint between global and local label correlations, and a label correlation guided multi-label supervised contrastive loss. RESULTS: To demonstrate the superior performance of our proposed approach, we conduct three times 5-fold cross-validation experiments on the CheXpert dataset. Our approach obtains an average F1 score of 44.6% and an AUC of 76.5%, achieving a 7.7% and 1.3% improvement compared to the state-of-the-art results. CONCLUSION: More accurate label correlations and full utilization of the learned label correlations help learn more discriminative label-level features. Experimental results demonstrate that our approach achieves exceptionally competitive performance compared to the state-of-the-art algorithms.

Estimation of heart inclination angle using posteroanterior chest radiograph and comparison with cardiac axis obtained from synthesized vectorcardiogram

ABSTRACT In medical practice, it is common to perform electrocardiography exams and by mathematical transformations to obtain the vectorcardiogram. The vectorcardiogram provides important information for medical diagnosis, such as the angle of inclination of the heart. This article aims to present a methodology for estimating the QRS vector-related angle of the heart using a posteroanterior chest radiograph image. We used an open source image processing software (Icy software version 2.3.0.0, Institut Pasteur, France, 2021) to perform a manual measurement of the target angle by analyzing relevant morphological structures from the x-ray images and using some functions to help the user to measure it. 18 radiographic images were selected to measure the angle of the heart by two independent individuals. The measured angles were compared using the mean absolute error (MAE). We then computed the QRS peak elevation angles of the vectorcardiogram (VCG) of the 57 patients collected at Dante Pazzanese Institute of Cardiology. In addition, an individual was randomly selected to measure a set of 57 radiographic images of these same patients. We performed the statistical treatments and the results suggested that the proposed manual method may be an alternative, viable and fast approach to estimating the anatomical heart axis for the purpose of aiding in medical diagnosis. However, further comparisons with more data and information are needed to determine its validity and possible method improvements.

HRCTCov19-a high-resolution chest CT scan image dataset for COVID-19 diagnosis and differentiation.

INTRODUCTION: Computed tomography (CT) was a widely used diagnostic technique for COVID-19 during the pandemic. High-Resolution Computed Tomography (HRCT), is a type of computed tomography that enhances image resolution through the utilization of advanced methods. Due to privacy concerns, publicly available COVID-19 CT image datasets are incredibly tough to come by, leading to it being challenging to research and create AI-powered COVID-19 diagnostic algorithms based on CT images. DATA DESCRIPTION: To address this issue, we created HRCTCov19, a new COVID-19 high-resolution chest CT scan image collection that includes not only COVID-19 cases of Ground Glass Opacity (GGO), Crazy Paving, and Air Space Consolidation but also CT images of cases with negative COVID-19. The HRCTCov19 dataset, which includes slice-level and patient-level labeling, has the potential to assist in COVID-19 research, in particular for diagnosis and a distinction using AI algorithms, machine learning, and deep learning methods. This dataset, which can be accessed through the web at http://databiox.com , includes 181,106 chest HRCT images from 395 patients labeled as GGO, Crazy Paving, Air Space Consolidation, and Negative.

Interpretable CNN-Multilevel Attention Transformer for Rapid Recognition of Pneumonia From Chest X-Ray Images.

Chest imaging plays an essential role in diagnosing and predicting patients with COVID-19 with evidence of worsening respiratory status. Many deep learning-based approaches for pneumonia recognition have been developed to enable computer-aided diagnosis. However, the long training and inference time makes them inflexible, and the lack of interpretability reduces their credibility in clinical medical practice. This paper aims to develop a pneumonia recognition framework with interpretability, which can understand the complex relationship between lung features and related diseases in chest X-ray (CXR) images to provide high-speed analytics support for medical practice. To reduce the computational complexity to accelerate the recognition process, a novel multi-level self-attention mechanism within Transformer has been proposed to accelerate convergence and emphasize the task-related feature regions. Moreover, a practical CXR image data augmentation has been adopted to address the scarcity of medical image data problems to boost the model's performance. The effectiveness of the proposed method has been demonstrated on the classic COVID-19 recognition task using the widespread pneumonia CXR image dataset. In addition, abundant ablation experiments validate the effectiveness and necessity of all of the components of the proposed method.

Thoracic cavity of the Shirazi cats: New insights using computed tomography and magnetic resonance imaging.

Our study provided a comprehensive characterization of the thorax of Shirazi cats by comparing the relevant soft and bone windows of computed tomography (CT) and magnetic resonance imaging (MRI) with cross, sagittal and coronal sectional anatomy. We outlined the mediastinum and its anatomic relationships with the trachea, oesophagus, lungs, heart, cranial and caudal vena cavae, and other thoracic structures using the data series gathered from adult normal Shirazi cats. The cranial mediastinum extended from the thoracic inlet to the 4th intercostal space, the middle mediastinum extended from the 5th and 7th intercostal spaces and was occupied by the heart and large blood vessels and the caudal mediastinum extended as a short and narrow portion from the 8th intercostal space to the diaphragm. The contents of the mediastinum and its relationship with the lungs and diaphragm were clearly presented in coronal-sectional anatomy and CT slices. The diaphragm was clearly observed in the lung windows of the ventral thorax. Sagittal-sectional anatomy and CT clarified the thorax's architecture and its contents, with higher density in the soft windows. The distribution of thoracic vessels on cross- and coronal-contrast CT scans was clearly visible. In addition, MRI scans provided an excellent anatomic reference of the thorax with the help of cross, coronal and sagittal scans, especially in the heart and blood vessels. Our study provides a valuable atlas for the diagnosis of malformations of the thoracic structures and offers better assessments for helping veterinary radiologists and clinicians in diagnostic processes.

Framework for dual-energy-like chest radiography image synthesis from single-energy computed tomography based on cycle-consistent generative adversarial network.

BACKGROUND: Dual-energy (DE) chest radiography (CXR) enables the selective imaging of two relevant materials, namely, soft tissue and bone structures, to better characterize various chest pathologies (i.e., lung nodule, bony lesions, etc.) and potentially improve CXR-based diagnosis. Recently, deep-learning-based image synthesis techniques have attracted considerable attention as alternatives to existing DE methods (i.e., dual-exposure-based and sandwich-detector-based methods) because software-based bone-only and bone-suppression images in CXR could be useful. PURPOSE: The objective of this study was to develop a new framework for DE-like CXR image synthesis from single-energy computed tomography (CT) based on a cycle-consistent generative adversarial network. METHODS: The core techniques of the proposed framework are divided into three categories: (1) data configuration from the generation of pseudo CXR from single energy CT, (2) learning of the developed network architecture using pseudo CXR and pseudo-DE imaging using a single-energy CT, and (3) inference of the trained network on real single-energy CXR. We performed a visual inspection and comparative evaluation using various metrics and introduced a figure of image quality (FIQ) to consider the effects of our framework on the spatial resolution and noise in terms of a single index through various test cases. RESULTS: Our results indicate that the proposed framework is effective and exhibits potential synthetic imaging ability for two relevant materials: soft tissue and bone structures. Its effectiveness was validated, and its ability to overcome the limitations associated with DE imaging techniques (e.g., increase in exposure dose owing to the requirement of two acquisitions, and emphasis on noise characteristics) via an artificial intelligence technique was presented. CONCLUSIONS: The developed framework addresses X-ray dose issues in the field of radiation imaging and enables pseudo-DE imaging with single exposure.

Insights on Using Time-of-Flight Camera for Recovering Cardiac Pulse From Chest Motion in Depth Videos.

In this article, we introduce a novel use of depth camera to extract cardiac pulse signal from human chest area, in which the depth information is obtained from a near infrared sensor using time-of-flight technology. We successfully isolate weak chest motion due to heartbeat by processing a sequence of depth images without raising privacy concern. We discuss motion sensitivity in depth video with examples from actuator simulation and human chest motion. Compared to other imaging modalities, the depth image intensity can be directly used for micromotion reconstruction. To deal with the challenges of recovering heartbeat from the chest area, we develop a set of coherent processing techniques to suppress the unwanted motion interference from breathing motion and involuntary body motion and eventually obtain clean cardiac pulse signal. We, thus, derive inter-beat-interval, showing high consistency to the contact photoplethysmography. Additionally, we develop a graphical interpretation of the most and the less pulsatile principal components in eigen space. For validation, we test our method on ten healthy human subjects with different resting heart rates. More importantly, we conduct a set of experiments to study the robustness and weakness of our methods, including extended range, multi-subject, thickness of clothes and generation to other measurement site.

The relationship between thorax computed tomography findings and prognosis in patients diagnosed with COVID-19.

OBJECTIVE: This study aimed to demonstrate the relationship between thorax computed tomography (CT) findings at the time of admission and prognosis using a semiquantitative CT severity scoring system in patients diagnosed with coronavirus disease 2019 (COVID-19) who tested positive for reverse transcriptase polymerase chain reaction (RT-PCR). PATIENTS AND METHODS: A total of 305 patients aged 18 years and older who were diagnosed with COVID-19 confirmed by RT-PCR and underwent thorax CT at the time of admission, were included in the study between March and July 2020. The demographic data of the patients, their presenting complaints at the time of admission, RT-PCR results, and thorax CT images were scanned retrospectively from electronic medical records. Lesions on thorax CT were evaluated for the presence of ground glass opacity, consolidation, and septal thickening and scoring. RESULTS: No significant relationship was found between mortality and CT score or other parameters. A significant relationship was found between admission to the intensive care unit and CT scoring (p=0.014), aortic diameter (p=0.032), chronic pulmonary disease (p=0.004), halo sign (p=0.031), mortality (p<0.001), fever (p=0.038), and dyspnea (p=0.031). A statistically significant difference was detected in the score parameter between discharged patients and intensive care unit patients who survived and those who died (p<0.001). In the parameter of the number of lobes, a statistically significant difference was found only between discharged patients and intensive care unit patients who survived (p=0.016). CONCLUSIONS: Thorax CT is an advisor for early diagnosis, treatment, and prognosis assessment of the disease. Semiquantitative CT severity scoring can provide valuable information about the prognosis of the patient.

Camera-based Respiratory Imaging for Thoracic Asymmetry in Thoracic Surgery Patients.

The current tool for assessing thoracic asymmetry of thoracic surgery patients is inappropriate for timely or frequent clinical routines due to its dependency on empirical physical examinations or specialized machines. This study investigates the camera-based respiratory imaging for screening thoracic asymmetry, in an intelligent and convenient way. The respiratory heatmaps are generated based on the respiratory magnitudes, phases and angles extracted from the chest video, and bilateral chest region of interest are compared statistically. Due to the variability of chest respiratory direction, spatial enhancement (SDR and SPCA) algorithms are proposed to magnify the respiratory energy. The proposed framework was validated in a clinical trial involving 31 patients, recorded by a smartphone camera. A high correlation was found between the camera measurements and patients' thoracic status in both the visual imaging and quantified indices. The respiratory imaging of camera shows a clear potential for assessing chest abnormalities of thoracic surgery patients.

A Patient-Specific correspondence model to track tumor location in thorax during radiation therapy.

PURPOSE: We present a patient-specific model to estimate tumor location in the thorax during radiation therapy using chest surface displacement as the surrogate signal. METHODS: Two types of data are used for model construction: Four-dimensional computed tomography (4D-CT) images of the patient and the displacement of two points on the patient's skin on the thoracic area. Principal component analysis is used to fit the correspondence model. This model incorporates the recorded surrogate signals during radiation delivery as input and delivers the 3D trajectory of the tumor as output. We evaluated the accuracy of the proposed model on a respiratory phantom and five lung cancer patients. RESULTS: For the respiratory phantom, the location of the center of the sphere during treatment was calculated in three directions: Left-Right (LR), Anterior-Posterior (AP) and, Superior-Inferior (SI). The error of localization was less than 1 mm in the LR and AP directions and less than 2 mm in the SI direction. The location of the tumor center for two of the patients, and the location of the apex of the diaphragm for the other three, was calculated in three directions. For all patients, the localization error in the LR and AP directions was less than 1.1 mm for two fractions and the maximum localization error in the SI direction was 6.4 mm. CONCLUSIONS: This work presents a feasibility study of utilizing surface displacement data to locate the tumor in the thorax during radiation treatment. Future work will validate the model on a larger patient population.

Effect of automated head-thorax elevation during chest compressions on lung ventilation: a model study.

Our goal was to investigate the effects of head-thorax elevation (HUP) during chest compressions (CC) on lung ventilation. A prospective study was performed on seven human cadavers. Chest was automatically compressed-decompressed in flat position and during progressive HUP from 18 to 35°. Lung ventilation was measured with electrical impedance tomography. In each cadaver, 5 sequences were randomly performed: one without CC at positive end-expiratory pressure (PEEP) 0cmH2O, 3 s with CC at PEEP0, 5 or 10cmH2O and 1 with CC and an impedance threshold device at PEEP0cmH2O. The minimal-to-maximal change in impedance (VTEIT in arbitrary unit a.u.) and the minimal impedance in every breathing cycle (EELI) the) were compared between flat, 18°, and 35° in each sequence by a mixed-effects model. Values are expressed as median (1st-3rd quartiles). With CC, between flat, 18° and 35° VTEIT decreased at each level of PEEP. It was 12416a.u. (10,689; 14,442), 11,239 (7667; 13,292), and 6457 (4631; 9516), respectively, at PEEP0. The same was true with the impedance threshold device. EELI/VTEIT significantly decreased from - 0.30 (- 0.40; - 0.15) before to - 1.13 (- 1.70; - 0.61) after the CC (P = 0.009). With HUP lung ventilation decreased with CC as compared to flat position. CC are associated with decreased in EELI.

[Comment on article "Correlation between ultrasonography and thorax portable radiography in patients with COVID-19"]. / Comentario al artículo "Correlación entre ultrasonografía y radiografía portátil de tórax en pacientes con COVID-19".

Point-of-care ultrasound is an emerging tool in critical care areas. In the study we are discussing, the ultrasonographic findings are compared and contrasted with the radiographic ones in patients with COVID-19.

El ultrasonido en el punto de atención es una herramienta emergente en la atención de las áreas críticas. En el estudio que comentamos, se comparan los hallazgos ultrasonográficos y se contrastan con los radiográficos en pacientes con COVID-19.