Objectives The study was to apply deep learning (DL) with convolutional neural networks (CNNs) to laryngoscopic imaging for assisting in real-time automated segmentation and classification of vocal cord leukoplakia. Methods This was a single-center retrospective diagnostic study included 216 patients who underwent laryngoscope and pathological examination from October 1, 2018 through October 1, 2019. Lesions were classified as nonsurgical group (NSG) and surgical group (SG) according to pathology. All selected images of vocal cord leukoplakia were annotated independently by 2 expert endoscopists and divided into a training set, a validation set, and a test set in a ratio of 6:2:2 for training the model. Results Among the 260 lesions identified in 216 patients, 2220 images from narrow band imaging (NBI) and 2144 images from white light imaging (WLI) were selected. For segmentation, the average intersection-over-union (IoU) value exceeded 70%. For classification, the model was able to classify the surgical group (SG) by laryngoscope with a sensitivity of 0.93 and specificity of 0.94 in WLI, and a sensitivity of 0.99 and specificity of 0.97 in NBI. Moreover, this model achieved a mean average precision (mAP) of 0.81 in WLI and 0.92 in NBI with an IoU> 0.5. Conclusions The study found that a model developed by applying DL with CNNs to laryngoscopic imaging results in high sensitivity, specificity, and mAP for automated segmentation and classification of vocal cord leukoplakia. This finding shows promise for the application of DL with CNNs in assisting in accurate diagnosis of vocal cord leukoplakia from WLI and NBI.

Objective: Persistent pulmonary interstitial emphysema (PPIE) is always related to mechanical ventilation and preterm. Its relationship with respiratory infection has rarely been reported in the literature. PPIE is difficult to diagnosis, always mimics with other cystic lesions. The objective of this study was to evaluate clinicopathological radiographic features of PPIE with respiratory infection and to detect the possible infectious pathogens. Methods: We retrospectively reviewed a total of 237 patients pathologically diagnosed with cystic lesions in West China Hospital of Sichuan University from January 2011 to April 2019. This retrospective cohort study analyzed clinicopathological radiographic features and to detect the infectious pathogens by metagenomic next-generation sequencing (mNGS). Results: Six cases were presented with primary syndrome of respiratory infection. There were four girls and two boys, ranged from 2 months to 5 years. 100% (5/5) available cases were full-term and without mechanical ventilation. CCAM were suspected in 66.7% (4/6) patients. 66.7% (4/6) cases affected only a single lobe, and 33.3% (2/6) cases affected both lung lobes. The pathologic characteristics showed lung cysts with variable size along the bronchovaslcular bundles, the cysts had a discontinuous fibrotic wall with a smooth inner surface, lined with uninucleated and/or multinucleated macrophages. Conclusions: Six rare cases of PPIE with respiratory infection were treated by lobectomy. All available five cases were full-term infants without mechanical ventilation. And we firstly tried to detect of infectious pathogens by mNGS, however, there was no certain infectious pathogen associated with PPIE in our study.