scholarly journals Artificial Intelligence and Democratization of the Use of Lung Ultrasound in COVID-19: On the Feasibility of Automatic Calculation of Lung Ultrasound Score

2022 ◽  
Vol 2 (1) ◽  
pp. 17-25
Author(s):  
Jorge Camacho ◽  
Mario Muñoz ◽  
Vicente Genovés ◽  
Joaquín L. Herraiz ◽  
Ignacio Ortega ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Real Time ◽  
Lung Ultrasound ◽  
Detection Algorithm ◽  
Image Processing Algorithm ◽  
Automatic Calculation ◽  
Ultrasound Score ◽  
Experienced Physician ◽  
Aided Diagnosis ◽  
B Lines

During the COVID-19 pandemic, lung ultrasound has been revealed as a powerful technique for diagnosis and follow-up of pneumonia, the principal complication of SARS-CoV-2 infection. Nevertheless, being a relatively new and unknown technique, the lack of trained personnel has limited its application worldwide. Computer-aided diagnosis could possibly help to reduce the learning curve for less experienced physicians, and to extend such a new technique such as lung ultrasound more quickly. This work presents the preliminary results of the ULTRACOV (Ultrasound in Coronavirus disease) study, aimed to explore the feasibility of a real-time image processing algorithm for automatic calculation of the lung ultrasound score (LUS). A total of 28 patients positive on COVID-19 were recruited and scanned in 12 thorax zones following the lung score protocol, saving a 3 s video at each probe position. Those videos were evaluated by an experienced physician and by a custom developed automated detection algorithm, looking for A-Lines, B-Lines, consolidations, and pleural effusions. The agreement between the findings of the expert and the algorithm was 88.0% for B-Lines, 93.4% for consolidations and 99.7% for pleural effusion detection, and 72.8% for the individual video score. The standard deviation of the patient lung score difference between the expert and the algorithm was ±2.2 points over 36. The exam average time with the ULTRACOV prototype was 5.3 min, while with a conventional scanner was 12.6 min. Conclusion: A good agreement between the algorithm output and an experienced physician was observed, which is a first step on the feasibility of developing a real-time aided-diagnosis lung ultrasound equipment. Additionally, the examination time was reduced to less than half with regard to a conventional ultrasound exam. Acquiring a complete lung ultrasound exam within a few minutes is possible using fairly simple ultrasound machines that are enhanced with artificial intelligence, such as the one we propose. This step is critical to democratize the use of lung ultrasound in these difficult times.

scholarly journals Lung ultrasound is useful in oncohematologic patients with respiratory dysfunction admitted to an Intensive Care Unit (ICU): a pilot study

2020 ◽  
Vol 22 (2) ◽  
pp. 1780182
Author(s):  
Cecilia Gomez Ravetti ◽  
Thiago Braganca Lana Silveira Ataide ◽  
Lidia Miranda Barreto ◽  
Fabricio De Lima Bastos ◽  
Angelica Gomide dos Reis Gomes ◽  
...  
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Pilot Study ◽  
Lung Ultrasound ◽  
Respiratory Dysfunction ◽  
Icu Admission ◽  
Immunosuppressed Patients ◽  
Show Difference ◽  
Ultrasound Score ◽  
B Lines

Aims: This pilot study aimed to evaluate the usefulness of a sequential lung ultrasound score (LUS) in immunosuppressed patients with oncohematologic diseases and acute respiratory dysfunction hospitalized in an intensive care unit (ICU).Materials and methods: LUS was calculated at ICU admission, after 24 h, 48 h and at discharge. A score ranging from 0 to 26 was attributed according to the number of B lines, presence of lung consolidation and pleural effusion.Results: Twenty-six patients were included. The median age was 50 years [interquartile range (IQR) 21] and 14 (54%) were male. LUS on the day of ICU admission was significantly higher in non-survivors compared to survivors (13 [5] vs 9 [9], respectively; p=0.047). The median delta LUS (LUS_D2 – LUS_D1) did not show difference between survivors and non-survivors (2 [0-7.5] vs 1 [-1.5 – 5], p=0.33). Among patients initially submitted to noninvasive mechanical ventilation (NIMV), no difference in LUS at inclusion or after 24 h was found between those who succeeded or failed on this support.Conclusion: The use of LUS to quantify lung aeration loss in oncohematologic patients hospitalized in an ICU due to acute respiratory dysfunction might be a helpful tool to predict the severity of the illness.

scholarly journals Point-of-Care Lung Ultrasound for COVID-19: Findings and Prognostic Implications From 105 Consecutive Patients

2021 ◽  
Vol 36 (3) ◽  
pp. 334-342 ◽  
Author(s):  
Kosuke Yasukawa ◽  
Taro Minami ◽  
David R. Boulware ◽  
Ayako Shimada ◽  
Ernest A. Fischer
Keyword(s):  
Negative Predictive Value ◽  
Predictive Value ◽  
Prognostic Value ◽  
Lung Ultrasound ◽  
Point Of Care ◽  
Respiratory Support ◽  
Point Of Care Ultrasound ◽  
Bilateral Involvement ◽  
Ultrasound Score ◽  
B Lines

Background: The prognostic value of point-of-care lung ultrasound has not been evaluated in a large cohort of patients with COVID-19 admitted to general medicine ward in the United States. The aim of this study was to describe lung ultrasound findings and their prognostic value in patients with COVID-19 admitted to internal medicine ward. Method: This prospective observational study consecutively enrolled 105 hospitalized participants with COVID-19 at 2 tertiary care centers. Ultrasound was performed in 12 lung zones within 24 hours of admission. Findings were assessed relative to 4 outcomes: intensive care unit (ICU) need, need for intensive respiratory support, length of stay, and death. Results: We detected abnormalities in 92% (97/105) of participants. The common findings were confluent B-lines (92%), non-homogenous pleural lines (78%), and consolidations (54%). Large confluent B-lines, consolidations, bilateral involvement, and any abnormality in ≥ 6 areas were associated with a longer hospitalization and need for intensive respiratory support. Large confluent B-lines and bilateral involvement were also associated with ICU stay. A total lung ultrasound score <5 had a negative predictive value of 100% for the need of intensive respiratory support. A higher total lung ultrasound score was associated with ICU need (median total 18 in the ICU group vs. 11 non-ICU, p = 0.004), a hospitalization ≥ 9d (15 vs 10, p = 0.016) and need for intensive respiratory support (18 vs. 8.5, P < 0.001). Conclusions: Most patients hospitalized with COVID-19 had lung ultrasound abnormalities on admission and a higher lung ultrasound score was associated with worse clinical outcomes except death. A low total lung ultrasound score (<5) had a negative predictive value of 100% for the need of intensive respiratory support. Point-of-care ultrasound can aid in the risk stratification for patients with COVID-19 admitted to general wards.

scholarly journals Severity Prediction System for Real Time Pothole Detection

2021 ◽  
Vol 23 (07) ◽  
pp. 1328-1334
Author(s):  
Sumit Bhimte ◽  
◽  
Hrishikesh hasabnis ◽  
Rohit Shirsath ◽  
Saurabh Sonar ◽  
...  
Keyword(s):  
Image Processing ◽  
Real Time ◽  
Detection System ◽  
Detection Algorithm ◽  
Prediction System ◽  
Image Processing Algorithm ◽  
Efficient Technology ◽  
Severity Prediction ◽  
Pothole Detection ◽  
Real Time Application

Pothole Detection System using Image Processing or using Accelerometer is not a new normal. But there is no real time application which utilizes both techniques to provide us with efficient solution. We present a system which can be useful for the drivers to determine the intensity of Pothole using both Image Processing Technology and Accelerometer device-based Algorithm. The challenge in building this system was to efficiently detect a Pothole present in roads, to analyze the severity of Pothole and to provide users with information like Road Quality and best possible route. We have used various algorithms for frequency-based pothole detection. We compared the results. Apart from that, we selected the best approach suitable for achieving the project goals. We have used a Simple Differentiation-based Edge Detection Algorithm for Image Processing. The system has been built on Map Interfaces for Android devices using Android Studio, which consists of usage of Image Processing Algorithm based Python frameworks which is a sub field of Machine Learning. It is backed by powerful DBMS. This project facilitates use of most efficient technology tools to provide a good user experience, real time application, reliability and improved efficiency.

scholarly journals B-line quantification: comparing learners novice to lung ultrasound assisted by machine artificial intelligence technology to expert review

2021 ◽  
Vol 13 (1) ◽  
Author(s):  
Frances M. Russell ◽  
Robert R. Ehrman ◽  
Allen Barton ◽  
Elisa Sarmiento ◽  
Jakob E. Ottenhoff ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Quantitative Analysis ◽  
Image Quality ◽  
Lung Ultrasound ◽  
Learning Algorithm ◽  
Intraclass Correlation ◽  
Training Session ◽  
Expert Review ◽  
Novice Learners ◽  
B Lines

Abstract Background The goal of this study was to assess the ability of machine artificial intelligence (AI) to quantitatively assess lung ultrasound (LUS) B-line presence using images obtained by learners novice to LUS in patients with acute heart failure (AHF), compared to expert interpretation. Methods This was a prospective, multicenter observational study conducted at two urban academic institutions. Learners novice to LUS completed a 30-min training session on lung image acquisition which included lecture and hands-on patient scanning. Learners independently acquired images on patients with suspected AHF. Automatic B-line quantification was obtained offline after completion of the study. Machine AI counted the maximum number of B-lines visualized during a clip. The criterion standard for B-line counts was semi-quantitative analysis by a blinded point-of-care LUS expert reviewer. Image quality was blindly determined by an expert reviewer. A second expert reviewer blindly determined B-line counts and image quality. Intraclass correlation was used to determine agreement between machine AI and expert, and expert to expert. Results Fifty-one novice learners completed 87 scans on 29 patients. We analyzed data from 611 lung zones. The overall intraclass correlation for agreement between novice learner images post-processed with AI technology and expert review was 0.56 (confidence interval [CI] 0.51–0.62), and 0.82 (CI 0.73–0.91) between experts. Median image quality was 4 (on a 5-point scale), and correlation between experts for quality assessment was 0.65 (CI 0.48–0.82). Conclusion After a short training session, novice learners were able to obtain high-quality images. When the AI deep learning algorithm was applied to those images, it quantified B-lines with moderate-to-fair correlation as compared to semi-quantitative analysis by expert review. This data shows promise, but further development is needed before widespread clinical use.

scholarly journals Correlation between Transthoracic Lung Ultrasound Score and HRCT Features in Patients with Interstitial Lung Diseases

2019 ◽  
Vol 8 (8) ◽  
pp. 1199 ◽  
Author(s):  
Milena Adina Man ◽  
Elena Dantes ◽  
Bianca Domokos Hancu ◽  
Cosmina Ioana Bondor ◽  
Alina Ruscovan ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Gold Standard ◽  
Lung Diseases ◽  
Lung Ultrasound ◽  
Interstitial Lung Diseases ◽  
High Resolution Computed Tomography ◽  
Complementary Role ◽  
Function Impairment ◽  
Ultrasound Score ◽  
B Lines

Chest high-resolution computed tomography (HRCT) is considered the “gold” standard radiological method in interstitial lung disease (ILD) patients. The objectives of our study were to evaluate the correlation between two transthoracic lung ultrasound (LUS) scores (total number of B-lines score = the total sum of B-lines in 10 predefined scanning sites and total number of positive chest areas score = intercostal spaces with ≥3 B-lines) and the features in HRCT simplified scores, in different interstitial disorders, between LUS scores and symptoms, as well as between LUS scores and pulmonary function impairment. We have evaluated 58 consecutive patients diagnosed with ILD. We demonstrated that there was a good correlation between the total number of B-lines score and the HRCT simplified score (r = 0.784, p < 0.001), and also a good correlation between the total number of positive chest areas score and the HRCT score (r = 0.805, p < 0.005). The results confirmed the value of using LUS as a diagnostic tool for the assessment of ILD compared to HRCT. The use of LUS in ILD patients can be a useful, cheap, accessible and radiation-free investigation and can play a complementary role in the diagnosis and monitoring of these patients.

scholarly journals Comprehensive Quantitative Assessment of Lung Liquid Clearance by Lung Ultrasound Score in Neonates with No Lung Disease during the First 24 Hours

2020 ◽  
Vol 2020 ◽  
pp. 1-5 ◽  
Author(s):  
Bin-Bin Guo ◽  
Kun-Kun Wang ◽  
Li Xie ◽  
Xiu-Juan Liu ◽  
Xiao-Ya Chen ◽  
...  
Keyword(s):  
Lung Disease ◽  
Chest Wall ◽  
Respiratory Symptoms ◽  
Quantitative Assessment ◽  
Lung Ultrasound ◽  
Ultrasound Score ◽  
Lung Liquid ◽  
B Lines ◽  
Anterior Posterior

Objectives. To comprehensively and quantitatively assess the process of lung liquid clearance using the lung ultrasound score. This study is to evaluate the whole healthy lungs of neonates during the first 24 h. Methods. Lung ultrasound was performed in neonates with no respiratory symptoms within 3 h after birth, and scans were then repeated at 6 hours and 24 hours, respectively. The entire chest wall was divided into 12 regions. The lung ultrasound scores of the anterior, posterior, upper, and lower regions and sum of all regions were calculated according to the ultrasound pattern of each region examined. Results. The total lung ultrasound score decreased gradually during the first 24 h, with the total lung ultrasound score at 6 h being significantly lower than that at <3 h (P<0.05). At <3 h, B-lines were more abundant in the posterior chest than in the anterior chest (P<0.001), and more B-lines were observed in the lower chest than in the upper chest (P<0.001). At 6 h and 24 h, there were no significant differences among the regions. Conclusion. Changes in the lung ultrasound score may quantitatively reflect the characteristics of different regions and processes of lung liquid clearance during the first 24 h.

Modified Lung Ultrasound Score for Assessing and Monitoring Pulmonary Aeration

2017 ◽  
Vol 38 (05) ◽  
pp. 530-537 ◽  
Author(s):  
Silvia Mongodi ◽  
Bélaïd Bouhemad ◽  
Anita Orlando ◽  
Andrea Stella ◽  
Guido Tavazzi ◽  
...  
Keyword(s):  
Lung Ultrasound ◽  
Intercostal Space ◽  
Icu Patients ◽  
Severe Loss ◽  
Constant Variance ◽  
Lung Aeration ◽  
Ultrasound Score ◽  
Space Width ◽  
B Lines ◽  
Pleural Involvement

Abstract Purpose Lung Ultrasound Score (LUSS) is a useful tool for lung aeration assessment but presents two theoretical limitations. First, standard LUSS is based on longitudinal scan and detection of number/coalescence of B lines. In the longitudinal scan pleura visualization is limited by intercostal space width. Moreover, coalescence of B lines to define severe loss of aeration is not suitable for non-homogeneous lung pathologies where focal coalescence is possible. We therefore compared longitudinal vs. transversal scan and also cLUSS (standard coalescence-based LUSS) vs. qLUSS (quantitative LUSS based on % of involved pleura). Materials and methods 38 ICU patients were examined in 12 thoracic areas in longitudinal and transversal scan. B lines (number, coalescence), subpleural consolidations (SP), pleural length and pleural involvement (> or ≤ 50 %) were assessed. cLUSS and qLUSS were computed in longitudinal and transversal scan. Results Transversal scan visualized wider (3.9 [IQR 3.8 – 3.9] vs 2.0 [1.6 – 2.5] cm, p < 0.0001) and more constant (variance 0.02 vs 0.34 cm, p < 0.0001) pleural length, more B lines (70 vs 59 % of scans, p < 0.0001), coalescence (39 vs 28 %, p < 0.0001) and SP (22 vs 14 %, p < 0.0001) compared to longitudinal scan. Pleural involvement > 50 % was observed in 17 % and coalescence in 33 % of cases. Focal coalescence accounted for 52 % of cases of coalescence. qLUSS-transv generated a different distribution of aeration scores compared to cLUSS-long (p < 0.0001). Conclusion In unselected ICU patients, variability of pleural length in longitudinal scans is high and focal coalescence is frequent. Transversal scan and quantification of pleural involvement are simple measures to overcome these limitations of LUSS.

scholarly journals Lung Ultrasound for Imaging of B-Lines in Dogs and Cats—A Prospective Study Investigating Agreement between Three Types of Transducers and the Accuracy in Diagnosing Cardiogenic Pulmonary Edema, Pneumonia and Lung Neoplasia

Animals ◽  
2021 ◽  
Vol 11 (11) ◽  
pp. 3279
Author(s):  
Andrzej Łobaczewski ◽  
Michał Czopowicz ◽  
Agata Moroz ◽  
Marcin Mickiewicz ◽  
Marta Stabińska ◽  
...  
Keyword(s):  
Pulmonary Edema ◽  
Lung Neoplasms ◽  
Lung Ultrasound ◽  
Companion Animals ◽  
Lung Neoplasm ◽  
Cardiogenic Pulmonary Edema ◽  
Area Under Roc Curve ◽  
Ultrasound Score ◽  
A Prospective Study ◽  
B Lines

Transthoracic heart and lung ultrasound (LUS) was performed in 200 dogs and cats with dyspnea to evaluate the agreement between the results obtained using three types of transducers (microconvex, linear, and phased array) and to determine the accuracy of LUS in discriminating between three conditions commonly causing dyspnea in companion animals: cardiogenic pulmonary edema (CPE), pneumonia, and lung neoplasm. The agreement beyond chance was assessed using the weighted Cohen’s kappa coefficient (κw). The highest values of κw (>0.9) were observed for the pair of microconvex and linear transducers. To quantify B-lines the lung ultrasound score (LUSscore) was developed as a sum of points describing the occurrence of B-lines for each of 8 standardized thoracic locations. The accuracy of LUSscore was determined using the area under ROC curve (AUROC). In dogs AUROC of LUSscore was 75.9% (CI 95%: 65.0% to 86.8%) for distinguishing between lung neoplasms and the two other causes of dyspnea. In cats AUROC of LUSscore was 83.6% (CI 95%: 75.2% to 92.0%) for distinguishing between CPE and the two other causes of dyspnea. The study shows that results obtained with microconvex and linear transducers are highly consistent and these two transducers can be used interchangeably. Moreover, the LUSscore may help identify dogs with lung neoplasms and cats with CPE, however its diagnostic accuracy is only fair to moderate.

scholarly journals BattleNet: Capturing Advantageous Battlefield in RTS Games (Student Abstract)

2020 ◽  
Vol 34 (10) ◽  
pp. 13849-13850
Author(s):  
Donghyeon Lee ◽  
Man-Je Kim ◽  
Chang Wook Ahn
Keyword(s):  
Artificial Intelligence ◽  
Decision Making ◽  
Real Time ◽  
Large Scale ◽  
Outcome Predictor ◽  
Short Term ◽  
Rts Game

In a real-time strategy (RTS) game, StarCraft II, players need to know the consequences before making a decision in combat. We propose a combat outcome predictor which utilizes terrain information as well as squad information. For training the model, we generated a StarCraft II combat dataset by simulating diverse and large-scale combat situations. The overall accuracy of our model was 89.7%. Our predictor can be integrated into the artificial intelligence agent for RTS games as a short-term decision-making module.

Sign in / Sign up

Export Citation Format

Share Document