scholarly journals Evaluation of Glucocorticoid Therapy in Asthma Children with Small Airway Obstruction Based on CT Features of Deep Learning

2021 ◽  
Vol 2021 ◽  
pp. 1-8
Author(s):  
Wei Zhang ◽  
Yang Wang
Keyword(s):  
Deep Learning ◽  
Airway Obstruction ◽  
Airway Inflammation ◽  
Maintenance Treatment ◽  
Learning Algorithm ◽  
Small Airway ◽  
Imaging Features ◽  
Ct Image ◽  
Deep Learning Algorithm ◽  
Small Airway Obstruction

This study was aimed at exploring the treatment of asthma children with small airway obstruction in CT imaging features of deep learning and glucocorticoid. A total of 145 patients meeting the requirements in hospital were included in this study, and they were randomly assigned to receive aerosolized glucocorticoid ( n = 45 ), aerosolized glucocorticoid combined with bronchodilator ( n = 50 ), or oral steroids ( n = 50 ) for 4 weeks after discharge. The lung function and fractional exhaled nitric oxide (FENO) indexes of the three groups were measured, respectively, and then the effective rates were compared to evaluate the clinical efficacy of glucocorticoids with different administration methods and combined medications in the short-term maintenance treatment after acute exacerbation of asthma. Deep learning algorithm was used for CT image segmentation. The CT image is sent to the workbench for processing on the workbench, and then the convolution operation is performed on each input pixel point during the image processing. After 4 weeks of maintenance treatment, FEF50 %, FEF75 %, and MMEF75/25 increased significantly, and FENO decreased significantly ( P < 0.01 ). The improvement results of FEF50 %, FEF75 %, MMEF75/25, and FENO after maintenance treatment were as follows: the oral hormone group was the most effective, followed by the combined atomization inhalation group, and the hormone atomization inhalation group was the least effective. The differences among them were statistically significant ( P < 0.05 ). The accuracy of artificial intelligence segmentation algorithm was 81%. All the hormones were more effective than local medication in the treatment of small airway function and airway inflammation. In the treatment of aerosol inhalation, the hormone combined with bronchiectasis drug was the most effective in improving small airway obstruction and reducing airway inflammation compared with single drug inhalation. Deep learning CT images are simple, noninvasive, and intuitively observe lung changes in asthma with small airway functional obstruction. Asthma with small airway functional obstruction has high clinical diagnosis and evaluation value.

scholarly journals Deep Learning Algorithms-Based CT Images in Glucocorticoid Therapy in Asthma Children with Small Airway Obstruction

2021 ◽  
Vol 2021 ◽  
pp. 1-12
Author(s):  
Yu Qin ◽  
Jing Wang ◽  
Yanjun Han ◽  
Ling Lu
Keyword(s):  
Deep Learning ◽  
Airway Obstruction ◽  
Small Airway ◽  
Ct Image ◽  
Asthmatic Children ◽  
Aerosol Inhalation ◽  
Group A ◽  
Group B ◽  
Small Airway Obstruction ◽  
Feno Level

CT image information data under deep learning algorithms was adopted to evaluate small airway function and analyze the clinical efficacy of different glucocorticoid administration ways in asthmatic children with small airway obstruction. The Res-NET in the deep learning algorithm was used to perform feature extraction, summary classification, and other reconstruction of CT images. A deep learning network model Mask-R-CNN was constructed to enhance the ability of image reconstruction. A total of 118 children hospitalized with acute exacerbation of asthma in the hospital were recruited. After acute exacerbation treatment, 96 children with asthma were screened out for small airway obstruction, which were divided into glucocorticoid aerosol inhalation group (group A, 32 cases), glucocorticoid combined with bronchodilator aerosol inhalation group (group B, 32 cases), and oral hormone therapy group (group C, 32 cases). Asthmatic children with small airway obstruction were screened after acute exacerbation treatment and were rolled into glucocorticoid aerosol inhalation group (group A), glucocorticoid combined with bronchodilators aerosol inhalation group (group B), and oral hormone therapy group (group C). Lung function indicators (maximal mid-expiratory flow (MMEF75 and 25), 50% forced expiratory flow (FEF50), and 75% forced expiratory flow (FEF75)), FeNO level, and airway inflammation indicators (IL-6, IL-35, and eosinophilic (EOS)) were compared before and one month after treatment. The ratio of airway wall thickness to outer diameter (T/D) and the percentage of airway wall area to total airway area (WA%) were measured by e-Health high-resolution CT (HRCT). The constructed network model was used to measure the patient's coronary artery plaque and blood vessel volume, and the image was reconstructed on the Res-Net network. It was found that the MSE value of the Res-Net network was the lowest, and the efficiency was very high during the training process. T/D and WA (%) of asthmatic children with small airway obstruction after treatment were significantly lower than those before treatment ( P < 0.01 ). After treatment, MMEF75/25 and FEF75 were significantly higher than those before treatment ( P < 0.05 ). Lung function-related indicator FEF50 was significantly higher than that before treatment ( P < 0.01 ). FeNO level after treatment was remarkably lower than that before treatment ( P < 0.01 ). In addition, lung function-related indicators, airway inflammation indicators, and FeNO level improved the most in group C, followed by group B, and those improvements in group A were the least obvious, with great differences among groups ( P < 0.05 ). In summary, the Res-Net model proposed was of certain feasibility and effectiveness for CT image segmentation and can effectively improve the clinical evaluation of patient CT image information. Glucocorticoids could improve small airway function and airway inflammation in asthmatic children with small airway obstruction, and oral corticosteroids were more effective than aerosol inhalation therapy.

scholarly journals Using deep learning to extract novel and quantitative imaging features from perfluoropropane contrast, sulphur hexafluoride contrast and non-contrast stress echocardiography images

2020 ◽  
Vol 41 (Supplement_2) ◽  
Author(s):  
C Dockerill ◽  
W Woodward ◽  
A McCourt ◽  
A Beqiri ◽  
A Parker ◽  
...  
Keyword(s):  
Deep Learning ◽  
Contrast Agents ◽  
Stress Echocardiography ◽  
Learning Algorithm ◽  
Dropout Rate ◽  
Imaging Features ◽  
Sulphur Hexafluoride ◽  
Deep Learning Algorithm ◽  
Feature Values ◽  
Contrast Studies

Abstract Background Stress echocardiography has become established as the most widely applied non-invasive imaging test for diagnosis of coronary artery disease within the UK. However, stress echocardiography has been substantially qualitative, rather than quantitative, based on visual wall motion assessment. For the first time, we have identified and validated quantitative descriptors of cardiac geometry and motion, extracted from ultrasound images acquired using contrast agents in an automated way. Purpose To establish whether these novel imaging features can be generated in an automated, quantifiable and reproducible way from images acquired with perfluoropropane contrast, as well as investigating how these extracted measures compare to those extracted from sulphur hexafluoride contrast and non-contrast studies. Methods 100 patients who received perfluoropropane contrast during their stress echocardiogram were recruited. Their stress echocardiography images were processed through a deep learning algorithm. Novel feature values were recorded and a subset of 10 studies were repeated. The automated measures of global longitudinal strain (GLS) and ejection fraction (EF) extracted from these images were compared to values previously extracted from sulphur hexafluoride contrast and non-contrast images using the same software. Results A full set of 31 novel imaging features were successfully extracted from 79 studies acquired using the perfluoropropane contrast agent with a dropout rate of 14% (n=92, 8 incomplete image sets). Repeated analysis in a subset of 10 perfluoropropane cases demonstrated excellent reproducibility of the extracted feature values (R2=1). Automated values of GLS and EF, at both rest (GLS = −16.4±4.8%, EF = 63±13%) and stress stages (GLS = −17.7±5.8%, EF = 68±11%), were extracted from 83 perfluoropropane studies, with a dropout rate of 16% (n=99, fewer incomplete sets as short axis view not required). The ranges of GLS and EF measures extracted from the perfluoropropane images were comparable to the other contrast studies (n=222) (Rest GLS = −16.8±5.8%, Rest EF = 63±10%; Stress GLS = −19.1±6.7%, Stress EF = 71±9%) and non-contrast studies (n=86) (Rest GLS = −15.7±5.3%, Rest EF = 57±10%; Stress GLS = −17.3±6.4%, Stress EF = 61±14%). Conclusions Novel features and clinically relevant measures were extracted from images acquired using perfluoropropane contrast for the first time in a fully automated and reproducible way using a deep learning algorithm. The analysis failure rate and generated measures are comparable to those extracted from images using other commonly used sulphur hexafluoride contrast agents and non-contrast stress echocardiography studies. These findings demonstrate that deep learning algorithms can be used for automated quantitative analysis of stress echocardiograms acquired using various contrast agents and in non-contrast studies to improve stress echocardiography practice. Funding Acknowledgement Type of funding source: Private company. Main funding source(s): Lantheus Medical Imaging, Inc.

scholarly journals TRACKING AND PREDICTING COVID-19 RADIOLOGICAL TRAJECTORY USING DEEP LEARNING ON CHEST X-RAYS: INITIAL ACCURACY TESTING

2020 ◽  
Author(s):  
S. Duchesne ◽  
D. Gourdeau ◽  
P. Archambault ◽  
C. Chartrand-Lefebvre ◽  
L. Dieumegarde ◽  
...  
Keyword(s):  
Deep Learning ◽  
Learning Algorithm ◽  
Roc Curves ◽  
Quantitative Information ◽  
Lung Lesion ◽  
Imaging Features ◽  
X Rays ◽  
Disease Trajectory ◽  
Deep Learning Algorithm ◽  
Deep Learning Features

ABSTRACTBackgroundDecision scores and ethically mindful algorithms are being established to adjudicate mechanical ventilation in the context of potential resources shortage due to the current onslaught of COVID-19 cases. There is a need for a reproducible and objective method to provide quantitative information for those scores.PurposeTowards this goal, we present a retrospective study testing the ability of a deep learning algorithm at extracting features from chest x-rays (CXR) to track and predict radiological evolution.Materials and MethodsWe trained a repurposed deep learning algorithm on the CheXnet open dataset (224,316 chest X-ray images of 65,240 unique patients) to extract features that mapped to radiological labels. We collected CXRs of COVID-19-positive patients from two open-source datasets (last accessed on April 9, 2020)(Italian Society for Medical and Interventional Radiology and MILA). Data collected form 60 pairs of sequential CXRs from 40 COVID patients (mean age ± standard deviation: 56 ± 13 years; 23 men, 10 women, seven not reported) and were categorized in three categories: “Worse”, “Stable”, or “Improved” on the basis of radiological evolution ascertained from images and reports. Receiver operating characteristic analyses, Mann-Whitney tests were performed.ResultsOn patients from the CheXnet dataset, the area under ROC curves ranged from 0.71 to 0.93 for seven imaging features and one diagnosis. Deep learning features between “Worse” and “Improved” outcome categories were significantly different for three radiological signs and one diagnostic (“Consolidation”, “Lung Lesion”, “Pleural effusion” and “Pneumonia”; all P < 0.05). Features from the first CXR of each pair could correctly predict the outcome category between “Worse” and “Improved” cases with 82.7% accuracy.ConclusionCXR deep learning features show promise for classifying the disease trajectory. Once validated in studies incorporating clinical data and with larger sample sizes, this information may be considered to inform triage decisions.

scholarly journals Is Intensity Inhomogeneity Correction Useful for Classification of Breast Cancer in Sonograms Using Deep Neural Network?

2018 ◽  
Vol 2018 ◽  
pp. 1-10 ◽  
Author(s):  
Chia-Yen Lee ◽  
Guan-Lin Chen ◽  
Zhong-Xuan Zhang ◽  
Yi-Hong Chou ◽  
Chih-Chung Hsu
Keyword(s):  
Deep Learning ◽  
Malignant Tumor ◽  
High Efficiency ◽  
Learning Algorithm ◽  
Intensity Inhomogeneity ◽  
Imaging Features ◽  
Learning Models ◽  
Deep Learning Algorithm ◽  
Inhomogeneity Correction ◽  
Public Data

The sonogram is currently an effective cancer screening and diagnosis way due to the convenience and harmlessness in humans. Traditionally, lesion boundary segmentation is first adopted and then classification is conducted, to reach the judgment of benign or malignant tumor. In addition, sonograms often contain much speckle noise and intensity inhomogeneity. This study proposes a novel benign or malignant tumor classification system, which comprises intensity inhomogeneity correction and stacked denoising autoencoder (SDAE), and it is suitable for small-size dataset. A classifier is established by extracting features in the multilayer training of SDAE; automatic analysis of imaging features by the deep learning algorithm is applied on image classification, thus allowing the system to have high efficiency and robust distinguishing. In this study, two kinds of dataset (private data and public data) are used for deep learning models training. For each dataset, two groups of test images are compared: the original images and the images after intensity inhomogeneity correction, respectively. The results show that when deep learning algorithm is applied on the sonograms after intensity inhomogeneity correction, there is a significant increase of the tumor distinguishing accuracy. This study demonstrated that it is important to use preprocessing to highlight the image features and further give these features for deep learning models. In this way, the classification accuracy will be better to just use the original images for deep learning.

scholarly journals Deep Learning-Based Identification of Spinal Metastasis in Lung Cancer Using Spectral CT Images

2021 ◽  
Vol 2021 ◽  
pp. 1-7
Author(s):  
Xiaojie Fan ◽  
Xiaoyu Zhang ◽  
Zibo Zhang ◽  
Yifang Jiang
Keyword(s):  
Lung Cancer ◽  
Deep Learning ◽  
Detection Rate ◽  
Learning Algorithm ◽  
Spinal Metastasis ◽  
Ct Images ◽  
Ct Image ◽  
Spectral Ct ◽  
Deep Learning Algorithm ◽  
Noise Ratio

In this study, deep learning algorithm-based energy/spectral computed tomography (CT) for the spinal metastasis from lung cancer was used. A dilated convolutional U-Net model (DC-U-Net model) was first proposed, which was used to segment the energy/spectral CT image of patients with the spinal metastasis from lung cancer. Subsequently, energy/spectral CT images under different energy levels were collected for the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) comparison. It was found the learning rate of the model decreased exponentially as the number of training increased, with the lung contour segmented out of the image. Under 40–65 keV, the CT value of bone metastasis from lung cancer decreased with increasing energy, as with the average rank sum test result. The SNR and CNR values were the highest under 60 keV. The detection rate of the deep learning algorithm below 60 keV was 81.41%, and that of professional doctors was 77.56%. The detection rate of the deep learning algorithm below 140 keV was 66.03%, and that of professional doctors was 64.74%. In conclusion, the DC-U-Net model demonstrates better segmentation effects versus the convolutional neutral networ k (CNN), with the lung contour segmented. Further, a higher energy level leads to worse segmentation effects on the energy/spectral CT image.

scholarly journals CT Image Feature Diagnosis on the Basis of Deep Learning Algorithm for Preoperative Patients and Complications of Transcatheter Aortic Valve Implantation

2021 ◽  
Vol 2021 ◽  
pp. 1-7
Author(s):  
Xiong Zheng ◽  
Zhang Qian ◽  
Xiaofang Wang ◽  
Zhen Zhang ◽  
Lei Liu
Keyword(s):  
Deep Learning ◽  
Aortic Valve ◽  
Learning Algorithm ◽  
Average Diameter ◽  
Ct Image ◽  
Transcatheter Aortic Valve ◽  
Deep Learning Algorithm ◽  
Aortic Valve Implantation ◽  
Valve Implantation ◽  
Short Diameter

This work was aimed to explore the role of CT angiography information provided by deep learning algorithm in the diagnosis and complications of the disease focusing on congenital aortic valve disease and severe aortic valve stenosis. 120 patients who underwent ultrasound cardiography for aortic stenosis and underwent transcatheter aortic valve implantation (TAVI) in hospital were selected as the research objects. Patients received CT examination of deep learning algorithm within one week. The measurement methods were long and short diameter method, area method, and perimeter method. The deep learning algorithm was used to measure the long and short diameter, area, and perimeter of the target area before CT image processing. The results showed that the average diameter of long and short diameter measurement was 95% CI (0.84, 0.92), the average diameter of perimeter measurement was 95% CI (0.68, 0.87), and the average diameter of area measurement was 95% CI (0.72, 0.91). Among the 52 patients, 35 cases were hypertension (67%), 13 cases were diabetes (25%), 6 cases were chronic renal insufficiency (Cr > 2 mg/dL) (11%) (2 cases were treated with hemodialysis, 3.8%), 11 patients had chronic pulmonary disease (21%), 9 patients had cerebrovascular disease (17.3%) and atrial flutter and atrial fibrillation. Deep learning can achieve excellent results in CT image processing, and it was of great significance for the diagnosis of TAVI patients, improving the success rate of treatment and the prognosis of patients.

scholarly journals Deep learning neural networks to differentiate Stafne’s bone cavity from pathological radiolucent lesions of the mandible in heterogeneous panoramic radiography

PLoS ONE ◽  
2021 ◽  
Vol 16 (7) ◽  
pp. e0254997
Author(s):  
Ari Lee ◽  
Min Su Kim ◽  
Sang-Sun Han ◽  
PooGyeon Park ◽  
Chena Lee ◽  
...  
Keyword(s):  
Deep Learning ◽  
High Performance ◽  
Imaging System ◽  
Panoramic Radiography ◽  
Learning Algorithm ◽  
Odontogenic Cysts ◽  
Surgical Removal ◽  
Imaging Features ◽  
Deep Learning Algorithm ◽  
Dentigerous Cysts

This study aimed to develop a high-performance deep learning algorithm to differentiate Stafne’s bone cavity (SBC) from cysts and tumors of the jaw based on images acquired from various panoramic radiographic systems. Data sets included 176 Stafne’s bone cavities and 282 odontogenic cysts and tumors of the mandible (98 dentigerous cysts, 91 odontogenic keratocysts, and 93 ameloblastomas) that required surgical removal. Panoramic radiographs were obtained using three different imaging systems. The trained model showed 99.25% accuracy, 98.08% sensitivity, and 100% specificity for SBC classification and resulted in one misclassified SBC case. The algorithm was approved to recognize the typical imaging features of SBC in panoramic radiography regardless of the imaging system when traced back with Grad-Cam and Guided Grad-Cam methods. The deep learning model for SBC differentiating from odontogenic cysts and tumors showed high performance with images obtained from multiple panoramic systems. The present algorithm is expected to be a useful tool for clinicians, as it diagnoses SBCs in panoramic radiography to prevent unnecessary examinations for patients. Additionally, it would provide support for clinicians to determine further examinations or referrals to surgeons for cases where even experts are unsure of diagnosis using panoramic radiography alone.

scholarly journals Deep Learning Algorithm Based on Analysis the Effect of Posterior Cervical Vertebral Canal Decompression Angioplasty in the Treatment of Ossification of Posterior Longitudinal Ligament of Cervical Spine by CT Image

2021 ◽  
Vol 37 (6-WIT) ◽  
Author(s):  
Yang Liu ◽  
Jianjun Kong
Keyword(s):  
Deep Learning ◽  
Cervical Spine ◽  
Spinal Stenosis ◽  
Learning Algorithm ◽  
Posterior Longitudinal Ligament ◽  
Ct Image ◽  
Hyperextension Injury ◽  
Vertebral Canal ◽  
Deep Learning Algorithm ◽  
Better Than

Original Article Deep learning algorithm based on analyzing the effect of posterior cervical vertebral canal decompression angioplastyin the treatment of ossification of posterior longitudinalligament of cervical spine by CT image Yang Liu1, Jianjun Kong2 ABSTRACTObjective: The paper uses the convolutional neural network algorithm in the deep learning algorithm to explore the therapeutic effect of surgical treatment of hyperextension injuries associated with ossification of the posterior longitudinal ligament of the cervical spine. Methods: In this retrospectively analyzed study 27 patients with hyperextension injury of the posterior longitudinal ligament of the cervical spine were selected from our hospital between August 2018 to July 2020. It included 21 males and 6 females; aged 36-79 years, with an average of 55.9 years. Results: Follow-up time of patients was 3-39 months, with an average of 17.4 months. The JOA score after surgery was significantly better than that before surgery (P﹤0.01), which was statistically significant; the improvement of JOA in patients undergoing anterior therapy was better than that in patients undergoing posterior therapy, which was statistically significant; the JOA improved in patients with minor violent injuries. The situation is significantly better than severe violent injuries, with statistical significance. The rate of postoperative JOA improvement was significantly correlated with the degree of nerve function retention of the injured spinal cord before surgery (P﹤0.01), and there was no significant correlation between the degree of spinal stenosis caused by ossification and the postoperative JOA improvement of patients.Conclusion: Convolutional neural network algorithm in the deep learning algorithm based on the cervical spine posterior longitudinal ligament ossification hyperextension injury was significantly improved after surgery. The less preoperative neurological damage, the postoperative neurological function, the degree of improvement, there was no significant correlation between the degree of spinal stenosis and the improvement of postoperative spinal cord function. For patients with ossification of the posterior longitudinal ligament, if there are neurological symptoms, early surgical treatment is recommended to relieve the compression, so as to prevent irreversible neurological damage caused by trauma. KEYWORDS: Cervical spine, Ossification of posterior longitudinal ligament, Hyperextension injury, Spinal stenosis rate, JOA improvement rate. doi: https://doi.org/10.12669/pjms.37.6-WIT.4857How to cite this:Liu Y, Kong J. Deep learning algorithm based on analyzing the effect of posterior cervical vertebral canal decompression angioplasty in the treatment of ossification of posterior longitudinal ligament of cervical spine by CT image. Pak J Med Sci. 2021;37(6):1630-1635. doi: https://doi.org/10.12669/pjms.37.6-WIT.4857 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Sign in / Sign up

Export Citation Format

Share Document