Respiratory Medicine, Journal Year: 2025, Volume and Issue: unknown, P. 108036 - 108036
Published: March 1, 2025
Language: Английский
BMJ Open, Journal Year: 2025, Volume and Issue: 15(1), P. e084311 - e084311
Published: Jan. 1, 2025
To identify the early predictors of a self-reported persistence long COVID syndrome (LCS) at 12 months after hospitalisation and to propose prognostic model its development. A combined cross-sectional prospective observational study. tertiary care hospital. 221 patients hospitalised for COVID-19 who have undergone comprehensive clinical, sonographic survey-based evaluation predischarge 1 month with subsequent 12-month follow-up. The final cohort included 166 had completed visit months. LCS discharge. Self-reported was detected in 76% participants 3 43% Patients reported incomplete recovery year were characterised by higher burden comorbidities (Charlson index 0.69±0.96 vs 0.31±0.51, p=0.001) residual pulmonary consolidations (1.56±1.78 0.98±1.56, p=0.034), worse blood pressure (BP) control (systolic BP 138.1±16.2 132.2±15.8 mm Hg, p=0.041), renal (estimated glomerular filtration rate 59.5±14.7 69.8±20.7 mL/min/1.73 m2, p=0.007) endothelial function (flow-mediated dilation brachial artery 10.4±5.4 12.4±5.6%, p=0.048), in-hospital levels liver enzymes (alanine aminotransferase (ALT) 76.3±60.8 46.3±25.3 IU/L, p=0.002) erythrocyte sedimentation (ESR) (34.3±12.1 28.3±12.6 mm/h, p=0.008), slightly indices ventricular longitudinal (left (LV) global strain (GLS) 18.0±2.4 17.0±2.3%, p=0011) Hospital Anxiety Depression Scale anxiety (7.3±4.2 5.6±3.8, p=0.011) depression scores (6.4±3.9 4.9±4.3, p=0.022) EFTER-COVID study physical symptoms score (12.3±3.8 9.2±4.2, p<0.001). At postdischarge, persisting differences marginally LV GLS, mitral E/e' ratio significantly both resting exertional versus complete recovery. Logistic regression machine learning-based binary classification models been developed predict Compared post-COVID-19 completely recovered hospital discharge, those subsequently 'very long' variety more pronounced abnormalities that mostly subsided month, except steady levels. simple artificial neural networks-based using peak ESR, creatinine, ALT weight loss during acute phase, 6-minute walk distance complex assessment as inputs has shown 92% accuracy an area under receiver-operator characteristic curve 0.931 prediction
Language: Английский
Citations
3
Emergency Radiology, Journal Year: 2025, Volume and Issue: unknown
Published: Feb. 14, 2025
Language: Английский
Citations
2
Ultrasonics, Journal Year: 2024, Volume and Issue: 140, P. 107251 - 107251
Published: Jan. 29, 2024
Lung ultrasound (LUS) has emerged as a safe and cost-effective modality for assessing lung health, particularly during the COVID-19 pandemic. However, interpreting LUS images remains challenging due to its reliance on artefacts, leading operator variability limiting practical uptake. To address this, we propose deep learning pipeline multi-class segmentation of objects (ribs, pleural line) artefacts (A-lines, B-lines, B-line confluence) in training phantom. Lightweight models achieved mean Dice Similarity Coefficient (DSC) 0.74, requiring fewer than 500 images. Applying this method real-time, at up 33.4 frames per second inference, allows enhanced visualisation these features This could be useful providing helping skill gap. Moreover, masks obtained from model enable development explainable measures disease severity, which have potential assist triage management patients. We suggest one such semi-quantitative measure called Artefact Score, is related percentage an intercostal space occupied by B-lines turn may associated with severity number conditions. show how transfer used train small datasets clinical images, identifying pathologies simple effusions consolidation DSC values 0.48 0.32 respectively. Finally, demonstrate DL translated into practice, implementing phantom alongside portable point-of-care system, facilitating bedside assessment improving accessibility LUS.
Language: Английский
Citations
11
Diagnostics, Journal Year: 2025, Volume and Issue: 15(1), P. 87 - 87
Published: Jan. 2, 2025
Background/Objectives: Point-of-care lung ultrasonography (LUS) represents an accurate diagnostic tool in older patients with respiratory failure. The integration of LUS ultrasonographic assessment diaphragm thickness and excursion, right vastus lateralis (RVL) muscle cross-sectional area (CSA) could provide real-time information on frailty sarcopenia. primary aim this proof-of-concept prospective study was to evaluate clinical correlates thoracic, diaphragmatic, muscular ultrasound characterize the associations between frailty, failure, sarcopenia hospitalized for acute complaints. Methods: Each 52 participants (age median 84, IQR 80–89 years old) underwent integrated LUS, RVL examination upon admission (T0) after 72 h hospitalization (T1). score used estimate interstitial syndrome severity. Diaphragm thickness, CSA were measured following a standardized protocol. Frailty assessed PC-FI (Primary Care-Frailty Index). Results: All exhibited multifactorial causes symptoms. T0 predicted 3-month rehospitalization. Frail higher scores T1. excursion reduced COPD heart failure those developing delirium during hospitalization. T1 negatively associated PC-FI. positive association obesity. Right T1, however, also Conclusions: Integrated lung, diaphragm, shows correlations several aspects that may help improve management geriatric illness.
Language: Английский
Citations
1
Animals, Journal Year: 2025, Volume and Issue: 15(1), P. 106 - 106
Published: Jan. 5, 2025
Thoracic point-of-care ultrasound (T-POCUS) has grown in popularity and usage small animal emergencies critical care settings due to its non-invasive nature, mobility, ability acquire images real time. This review summarizes current understanding about T-POCUS dogs cats with respiratory illnesses, including normal thoracic ultrasonography appearance numerous pathological situations. The basics of are covered, equipment, scanning procedures, picture settings. Practical applications patients distress discussed, an emphasis on pleural space abnormalities lung diseases. Ultrasound results define pulmonary disorders such as pneumonia, atelectasis, cardiogenic non-cardiogenic edema, lobe torsion, fibrosis, thromboembolism, neoplasms, bleeding. evaluation focuses diagnostic skills a variety clinical Limitations the need for more study standardize techniques, establish agreed terminology, create specialized educational routes highlighted.
Language: Английский
Citations
1
JAMA Cardiology, Journal Year: 2025, Volume and Issue: unknown
Published: Jan. 15, 2025
Lung ultrasound (LUS) aids in the diagnosis of patients with dyspnea, including those cardiogenic pulmonary edema, but requires technical proficiency for image acquisition. Previous research has demonstrated effectiveness artificial intelligence (AI) guiding novice users to acquire high-quality cardiac images, suggesting its potential broader use LUS. To evaluate ability AI guide acquisition diagnostic-quality LUS images by trained health care professionals (THCPs). In this multicenter diagnostic validation study conducted between July 2023 and December 2023, participants aged 21 years or older shortness breath recruited from 4 clinical sites underwent 2 examinations: 1 examination a THCP operator using Guidance other expert without AI. The THCPs (including medical assistants, respiratory therapists, nurses) standardized training before participation. software uses deep learning algorithms B-line annotation. Using an 8-zone protocol, automatically captures quality. primary end point was proportion THCP-acquired examinations quality according panel 5 masked readers, who provided remote review ground truth validation. intention-to-treat analysis included 176 (81 female [46.0%]; mean [SD] age, 63 [14] years; body mass index, 31 [8]). Overall, 98.3% (95% CI, 95.1%-99.4%) studies were quality, no statistically significant difference compared expert-acquired (difference, 1.7%; 95% -1.6% 5.0%). study, assistance achieved meeting standards experts This technology could extend access underserved areas lacking personnel. ClinicalTrials.gov Identifier: NCT05992324.
Language: Английский
Citations
1
Animals, Journal Year: 2025, Volume and Issue: 15(4), P. 549 - 549
Published: Feb. 13, 2025
Human literature describes vascular patterns in various types of lung consolidations; however, these changes have not been analyzed dogs and cats. This retrospective analysis medical records aimed to describe observed the airless subpleural tissue cats compare consolidations clinical diagnoses according parenchymal criteria described human literature. study included 347 634 dogs. was a encompassing obtained between 2018 2023. Lung ultrasound performed cases with different sonographically identified were selected. Airless categorized into five consolidations: shred, nodule, wedge sign, mass, sign. Further classification based on criteria, presence or absence bronchograms within regions. Bronchograms classified as air (dynamic and/or static), fluid bronchograms, mixed bronchograms. Vascular tree-like, residual, chaotic, "vascular sign", whether they continuous extended from chest wall not. It is possible identify characterize characteristics.
Language: Английский
Citations
1
BMC Medicine, Journal Year: 2025, Volume and Issue: 23(1)
Published: Feb. 24, 2025
Language: Английский
Citations
1
Breathe, Journal Year: 2025, Volume and Issue: 21(2), P. 240170 - 240170
Published: April 1, 2025
B-lines and pleural line thickening on LUS are sensitive but nonspecific signs of ILD. aids in early detection monitoring, HRCT PFT remain the gold standards. Limitations include operator dependence lack standardised protocols. https://bit.ly/41vUQSn.
Language: Английский
Citations
1
Ultrasonics, Journal Year: 2023, Volume and Issue: 132, P. 106994 - 106994
Published: March 30, 2023
Automated ultrasound imaging assessment of the effect CoronaVirus disease 2019 (COVID-19) on lungs has been investigated in various studies using artificial intelligence-based (AI) methods. However, an extensive analysis state-of-the-art Convolutional Neural Network-based (CNN) models for frame-level scoring, a comparative aggregation techniques video-level together with thorough evaluation capability these methodologies to provide clinically valuable prognostic-level score is yet missing within literature. In addition that, impact posterior probability assigned by network predicted frames as well temporal downsampling LUS data are topics not extensively investigated. This paper takes challenges providing benchmark methods from frame prognostic level. For deep learning evaluated additional best performing model transfer-learning settings. A novel cross-correlation based technique proposed video and exam-level scoring. Results showed that ResNet-18, when trained scratch, outperformed existing F1-Score 0.659. The method resulted 59.51%, 63.29%, 84.90% agreement clinicians at video, exam, levels, respectively; thus, demonstrating improved performances over state art. It was also found filtering shows higher comparison downsampling. All were conducted largest standardized validated dataset COVID-19 patients.
Language: Английский
Citations
23