Measuring performance in the interpretation of chest radiographs: a pilot study

Diagnostic performance of chest radiography in high COVID-19 prevalence setting: experience from a European reference hospital

Purpose

The study’s aim is to analyse the diagnostic performance of chest radiography (CXR) in patients with suspected coronavirus disease 19 (COVID-19).

Methods

We retrospectively considered 826 consecutive patients with suspected COVID-19 presenting to our emergency department (ED) from February 21 to March 31, 2020, in a high disease prevalence setting. We enrolled patients who underwent CXR and rhino-oropharyngeal swab for real-time reverse transcription-polymerase chain reaction (rRT-PCR). CXRs were evaluated by an expert radiologist; a second independent analysis was performed by two residents in consensus. All readers, blinded to rRT-PCR results, classified CXRs positive/negative depending on presence/absence of typical findings of COVID-19, using rRT-PCR as reference standard.

Results

We finally analysed 680 patients (median age 58); 547 (80%) tested positive for COVID-19. The diagnostic performance of CXR, interpreted by the expert reader, was as follows: sensitivity (79.0%; 95% CI: 75.3–82.3), specificity (81.2%; 95% CI: 73.5–87.5), PPV (94.5%;95% CI: 92.0–96.4), NPV (48.4%; 95% CI: 41.7–55.2), and accuracy (79.3%; 95% CI: 76.0–82.2). For the residents: sensitivity (75.1%; 95% CI: 71.2–78.7), specificity (57.9%; 95% CI: 49.9–66.4), PPV (88.0%; 95% CI: 84.7–90.8), NPV (36.2%; 95% CI: 29.7–43.0), and accuracy (71.6%; 95% CI: 68.1–75.0). We found a significant difference between the reporting sensitivity (p = 0.013) and specificity (p < 0.0001) of expert radiologist vs residents. CXR sensitivity was higher in patients with symptom onset > 5 days before ED presentation compared to ≤ 5 days (84.4% vs 70.7%).

Conclusions

CXR showed a sensitivity of 79% and a specificity of 81% in diagnosing viral pneumonia in symptomatic patients with clinical suspicion of COVID-19. Further studies in lower prevalence settings are needed.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10140-021-01946-x.

Amide Proton Transfer Imaging vs Diffusion Kurtosis Imaging for Predicting Histological Grade of Hepatocellular Carcinoma

Background

To investigate the value of amide proton transfer (APT) imaging in predicting the histological grade of hepatocellular carcinoma (HCC), compared with diffusion kurtosis imaging (DKI).

Methods

A total of 88 patients with HCC were enrolled and divided into four groups (G1, G2, G3, and G4) based on histologic grades. Preoperative APT signal intensity (SI), mean diffusivity (MD), mean kurtosis (MK) of HCC were measured and compared. Those quantitative magnetic resonance imaging (qMRI) parameters were compared using an analysis of variance. The correlations between the qMRI parameters and the histological grades were determined using Spearman's rank analysis. In addition, the predictive performance for differentiating low- (G1 and G2) from high-grade (G3 and G4) HCC was evaluated using receiver operating characteristic (ROC) curve analysis.

Results

Significant differences were found in APT SIs, MD, and MK among the four groups (P<0.05). Moderate to good relationships were found between the histologic grade of HCC and APT SI and MK (r=0.679, P<0.001 and r=0.539, P<0.001, respectively). The area under the ROC curves (AUCs) of APT SI, MK, and MD for differentiating low- from high-grade HCC were 0.890 (95%CI: 0.805–0.947), 0.765 (95%CI: 0.662–0.849) and 0.717 (95%CI: 0.611–0.808), respectively. Comparison of ROC curves showed a significantly higher AUC of APT SI compared with those of the DKI-derived parameters (P <0.05).

Conclusion

The APT imaging may be more accurate than DKI for predicting the histological grade of HCC.

Measuring the Effects of Education in Detecting Lung Cancer on Chest Radiographs: Utilization of a New Assessment Tool

This study was designed to evaluate the effect of group and individualized educational lectures to accurately interpret chest radiographs of lung cancer patients and to introduce a new educational tool in evaluating skills for reading chest radiographs. Utilizing "hotspot" technology will be instrumental in measuring the effect of education in interpreting chest radiographs. There were 48 participants in the study. Chest radiographs of 100 lung cancer patients and 11 healthy patients taken at various time points were used for evaluation. Using "hotspot" technology, lesions on each radiograph were outlined. Values were taken at baseline, after which the group received lectures. Several days later, they underwent exam 2. Exam 3 was conducted after individualized lectures. A final exam was taken after the participants underwent individualized training within 2 months. Scores significantly improved after the individual lessons (p < 0.001). This improvement in performance decreased in the final examination. Statistically significant differences were observed between exam 2 vs. exam 3 and exam 3 vs. the final exam (p < 0.001, p < 0.001). Participants demonstrated more improvement in detecting lesions in abnormal chest radiographs than in identifying normal ones. Although there was significant improvement in detecting abnormal radiographs by the end of the study (p < 0.001), no improvement was observed in detecting normal ones. We measured lung cancer detection rate using a new "hotspot" detection tool for chest radiographs. With the proposed scoring system, this tool could be objectively used in evaluating the educational effects.