Correlation between clinical features, high-resolution computed tomography findings, and a visual scoring system in patients with pneumonia due to Mycoplasma pneumoniae

The Diagnostic Value of High-Resolution Computed Tomography Features Combined with Mycoplasma Pneumoniae Ribonucleic Acid Load Detection for Refractory Mycoplasma Pneumonia

The aim of this study was to investigate the value of high-resolution computed tomography (CT) images and mycoplasma pneumoniae (MP) ribonucleic acid (RNA) load detection in the early diagnosis of refractory mycoplasma pneumoniae (RMP) and provide more methods for the diagnosis and treatment of RMP. Seventy children with MP were divided into the RMP group (H1 group, 31 cases) and the MP group (H2 group, 39 cases) according to pathological findings, and all of them underwent CT scanning. MP-RNA load and genotype distribution were analyzed in both groups, and the diagnostic efficacy of CT combined with MP-RNA load for RMP was calculated. The sensitivity of children in the H1 group to erythromycin (59.17% vs 71.56%) and clarithromycin (53.21% vs 67.03%) was lower than that in the H2 group, and the resistance rate of children in the H1 group to erythromycin (71.43% vs 67.53%) and clarithromycin (64.24% vs 50.37%) was higher than that in the H2 group (P < 0.05); the regression coefficients between lactate dehydrogenase (LDH) and the MPLI value of RMP were −0.064 and −0.413, respectively, which were significantly negatively correlated (P < 0.05); the accuracy (96.5%), sensitivity (92.5%), and specificity (88%) of CT + MP-RNA in the diagnosis of RMP were significantly higher than those of CT alone (91%, 88%, and 82%) and MP-RNA alone (88%, 84.5%, and 74%), which were significantly different (P < 0.05). The results of high MP-RNA load detection can be used as an indicator to predict RMP, and the diagnostic efficacy is significantly improved after combination with high-resolution CT, with high clinical application value.

Combined Radiographic Features and Age Can Distinguish Mycoplasma pneumoniae Pneumonia from Other Bacterial Pneumonias: Analysis Using the 16S rRNA Gene Sequencing Data

The study objective was to evaluate chest radiographic features that distinguish Mycoplasma pneumoniae pneumonia (MPP) from other bacterial pneumonias diagnosed based on the bacterial floral analysis with 16S rRNA gene sequencing, using bronchoalveolar lavage fluid samples directly obtained from pneumonia lesions. Patients were grouped according to the dominant bacterial phenotype; among 120 enrolled patients with CAP, chest CT findings were evaluated in 55 patients diagnosed with a mono-bacterial infection (one bacterial phylotype occupies more than 80% of all phylotypes in a sample) by three authorized respiratory physicians. Among this relatively small sample size of 55 patients with CAP, 10 had MPP, and 45 had other bacterial pneumonia and were categorized into four groups according to their predominant bacterial phylotypes. We created a new scoring system to discriminate MPP from other pneumonias, using a combination of significant CT findings that were observed in the M. pneumoniae group, and age (<60 years) (MPP−CTA scoring system). When the cutoff value was set to 1, this scoring system had a sensitivity of 80%, a specificity of 93%, a positive predictive value of 73%, and a negative predictive value of 95%. Among the CT findings, centrilobular nodules were characteristic findings in patients with MPP, and a combination of chest CT findings and age might distinguish MPP from other bacterial pneumonias.

Deep diagnostic agent forest (DDAF): A deep learning pathogen recognition system for pneumonia based on CT

BACKGROUND

Even though antibiotics agents are widely used, pneumonia is still one of the most common causes of death around the world. Some severe, fast-spreading pneumonia can even cause huge influence on global economy and life security. In order to give optimal medication regimens and prevent infectious pneumonia's spreading, recognition of pathogens is important.

METHOD

In this single-institution retrospective study, 2,353 patients with their CT volumes are included, each of whom was infected by one of 12 known kinds of pathogens. We propose Deep Diagnostic Agent Forest (DDAF) to recognize the pathogen of a patient based on ones' CT volume, which is a challenging multiclass classification problem, with large intraclass variations and small interclass variations and very imbalanced data.

RESULTS

The model achieves 0.899 ± 0.004 multi-way area under curves of receiver (AUC) for level-I pathogen recognition, which are five rough groups of pathogens, and 0.851 ± 0.003 AUC for level-II recognition, which are 12 fine-level pathogens. The model also outperforms the average result of seven human readers in level-I recognition and outperforms all readers in level-II recognition, who can only reach an average result of 7.71 ± 4.10% accuracy.

CONCLUSION

Deep learning model can help in recognition pathogens using CTs only, which might help accelerate the process of etiological diagnosis.

Application of CT images based on the optimal atlas segmentation algorithm in the clinical diagnosis of Mycoplasma Pneumoniae Pneumonia in Children

Objective

Use of optimal Atlas segmentation algorithm to study the imaging signs of mycoplasma pneumonia with multi-slice spiral CT (HRCT), and to explore the value of HRCT in the diagnosis and efficacy in evaluation of mycoplasma pneumonia in children.

Methods

The study retrospectively analyzed 72 patients diagnosed with mycoplasma pneumonia in our hospital from January 2017 to January 2019. The imaging data and clinical data of 72 patients were collected. The optimal Atlas segmentation algorithm was used to analyze the characteristics of CT examination, and the value of CT in the diagnosis of mycoplasma pneumonia and the evaluation of curative effect was summarized.

Results

Among all patients, 37 cases were unilateral lesions, 35 cases were bilateral lesions, 19 cases were in the left upper lobe, 24 cases were in the left lower lobe, 21 cases were in the right upper lobe, 13 cases were in the right middle lobe, 25 The lesion was located in the right lower lobe. The main CT findings of the lesions before treatment were large patchy, spot-shaped shadows, and strip-shaped or ground-glass shadows. After treatment, the main CT findings of the lesions were reduced lesion density and reduced lesion range.

Conclusion

CT can clearly show the pulmonary lesions of mycoplasma pneumonia, and its unique imaging signs can improve the clinical diagnosis accuracy. In addition, CT scans can evaluate the treatment effect according to the changes in the characteristics of the lesion, which has important value for the evaluation of the effect for clinical diagnosis and efficacy evaluation of mycoplasma pneumonia.

Preoperative serum lactate dehydrogenase level predicts progression and prognosis in patients with glioma

BACKGROUND

To evaluate the value of serum Lactate Dehydrogenase (LDH) level in predicting recurrence and the overall survival (OS) of glioma patients.

MATERIALS AND METHODS

A total number of 216 patients with glioma in our institution were retrospectively recruited to analyze the relationship between preoperative serum LDH level and prognosis.

RESULTS

Overall, the median age of patients was 46.0 (31.0-57.0) years old; 53.7% (116 of 216) of the enrolled patients were male. Multivariate analysis revealed that serum LDH level (odds ratio [OR] = 0.97, 95% confidence interval [CI] = 0.96-0.98, P < 0.001) and World Health Organization (WHO) grade (grade II: OR = 19.64, 95%CI = 5.56-69.35, P < 0.001; grade III: OR =1 9.50, 95%CI = 7.08-53.73, P < 0.001; grade IV: OR = 15.23, 95%CI = 4.94-46.97, P < 0.001) were significant and independent of 1-year Progression-free survival (PFS) after adjusting for confounders. The predictive performance of serum LDH level was represented with area under curve (AUC) = 0.741, 95%CI = 0.677-0.798. Multivariate Cox analysis revealed that LDH level (hazard ratio [HR] = 2.56, 95%CI = 1.59-4.15, P < 0.001) and WHO grade (grade II: HR = 4.58, 95%CI = 0.56-37.23, P = 0.155; grade III: HR = 16.35, 95%CI = 2.16-123.80, P = 0.007; grade IV: HR = 42.13, 95%CI = 5.83-304.47, P < 0.001) remained associated with survival at 2-year follow-up. At 3-year follow-up, lymphocyte count (HR = 0.68, 95%CI = 0.51-0.91, P = 0.008), LDH level (HR = 2.21, 95%CI = 1.40-3.49, P = 0.001), and WHO grade (grade II: HR = 1.44, 95%CI = 0.44-4.68, P = 0.543; grade III: HR = 4.99, 95%CI = 1.68-14.87, P = 0.004; grade IV: HR = 16.96, 95%CI = 6.13-46.93, P < 0.001) remained associated with survival in multivariate Cox analysis.

CONCLUSION

Our study demonstrated that preoperative serum LDH level could serve as a reliable indicator for predicting prognosis of glioma patients. Further multicenter studies are still required to verify our findings.

LDH, CRP and ALB predict nucleic acid turn negative within 14 days in symptomatic patients with COVID-19

AIMS

To search for biochemical indicators that can identify symptomatic patients with COVID-19 whose nucleic acid could turn negative within 14 days, and assess the prognostic value of these biochemical indicators in patients with COVID-19.

PATIENTS AND METHODS

We collected the clinical data of patients with COVID-19 admitted to our hospital, by using logistic regression analysis and AUC curves, explored the relationship between biochemical indicators and nucleic acid positive duration, the severity of COVID-19, and hospital stay respectively.

RESULTS

A total of two hundred and thirty-three patients with COVID-19 were enrolled in the study. We found patients whose nucleic acid turned negative within 14 days had lower LDH, CRP and higher ALB (P < 0.05). ROC curve results indicated that lower LDH, TP, CRP and higher ALB predicted the nucleic acid of patients turned negative within 14 days with statistical significance(P < 0.05), AST, LDH, CRP and PCT predicted the severe COVID-19 with statistical significance, and CRP predicted hospital stay >31days with statistical significance (P < 0.05). After verification, the probability of nucleic acid turning negative within 14 days in patients with low LDH (<256 U/L), CRP (<44.5 mg/L) and high ALB (>35.8 g/L) was about 4 times higher than that in patients with high LDH, CRP and low ALB (P < 0.05).

CONCLUSIONS

LDH, CRP and ALB are useful prognostic marker for predicting nucleic acid turn negative within 14 days in symptomatic patients with COVID-19.

High serum lactate dehydrogenase and dyspnea: Positive predictors of adverse outcome in critical COVID-19 patients in Yichang

BACKGROUND

Coronavirus disease 2019 (COVID-19) caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) outbreak in China, constitutes a Public Health Emergency of International Concern. It is well known that COVID-19 patients may have increased serum lactate dehydrogenase (LDH) levels in the early stage. The clinical changes in LDH may have predictive value in disease evolution and prognosis in critically ill COVID-19 patients.

AIM

To examine serum LDH and clinical characteristics in patients with COVID-19 and their predictive value for prognosis.

METHODS

This retrospective study analyzed the clinical data of forty-seven critical COVID-19 patients in the intensive care unit of the Third People's Hospital of Yichang City from January 27 to March 25, 2020 and divided them into survivors and non-survivors. The patients were diagnosed according to the World Health Organization interim guidance and critical cases met any one of the following criteria: Respiratory failure and required mechanical ventilation, the occurrence of shock, and the combined failure of other organs that required intensive care unit monitoring and treatments, according to the diagnostic criteria of critical COVID-19. Clinical data including symptoms, detection of SARS-CoV-2, chest computed tomography (CT) images, changes in serum LDH in different clinical phases, and prognosis were collected. Statistical analysis of the data was performed. Continuous variables were expressed as median (interquartile range) and compared with the Mann-Whitney U test. Categorical variables were compared with the Chi-square test. Survival data were analyzed using Kaplan-Meier survival curves and log-rank tests.

RESULTS

According to chest CT images, we observed the alveolitis and fibrosis stages in all critical patients in this study. Most non-survivors died in the fibrosis stage. Non-survivors had fewer days of hospitalization, shorter disease duration, shorter duration of alveolitis and fibrosis, and had dyspnea symptoms at disease onset (P = 0.05). Both first and lowest LDH values in the alveolitis stage were more pronounced in non-survivors than in survivors (449.0 U/L vs 288.0 U/L, P = 0.0243; 445.0 U/L vs 288.0 U/L, P = 0.0199, respectively), while the first, lowest and highest values of serum LDH in non-survivors were all significantly increased compared to survivors in the fibrosis phase (449.0 U/L vs 225.5 U/L, P = 0.0028; 432.0 U/L vs 191.0 U/L, P = 0.0007; 1303.0 U/L vs 263.5 U/L, P = 0.0001, respectively). The cut-off points of first LDH values in the alveolitis and fibrosis phase for distinction of non-survivors from survivors were 397.0 U/L and 263.0 U/L, respectively. In the fibrosis stage, non-survivors had more days with high LDH than survivors (7.0 d vs 0.0 d, P = 0.0002). Importantly, patients with high LDH had a significantly shorter median survival time than patients with low LDH in the alveolitis phase (22.0 d vs 36.5 d, P = 0.0002), while patients with high LDH also had a significantly shorter median survival time than patients with low LDH in the fibrosis phase (27.5 d vs 40.0 d, P = 0.0008). The proportion of non-survivors with detectable SARS-CoV-2 until death in the alveolitis stage was significantly increased compared with that in the fibrosis stage (100% vs 35.7%, P = 0.0220).

CONCLUSION

High LDH and dyspnea symptoms were positive predictors of an adverse outcome in critical COVID-19. The rapid progressive fibrosis stage was more perilous than the alveolitis stage, even if SARS-CoV-2 is undetectable.