Clinical screening of Nocardia in sputum smears based on neural networks.

Objective: Nocardia is clinically rare but highly pathogenic in clinical practice. Due to the lack of Nocardia screening methods, Nocardia is often missed in diagnosis, leading to worsening the condition. Therefore, this paper proposes a Nocardia screening method based on neural networks, aiming at quick Nocardia detection in sputum specimens with low costs and thereby reducing the missed diagnosis rate. Methods: Firstly, sputum specimens were collected from patients who were infected with Nocardia, and a part of the specimens were mixed with new sputum specimens from patients without Nocardia infection to enhance the data diversity. Secondly, the specimens were converted into smears with Gram staining. Images were captured under a microscope and subsequently annotated by experts, creating two datasets. Thirdly, each dataset was divided into three subsets: the training set, the validation set and the test set. The training and validation sets were used for training networks, while the test set was used for evaluating the effeteness of the trained networks. Finally, a neural network model was trained on this dataset, with an image of Gram-stained sputum smear as input, this model determines the presence and locations of Nocardia instances within the image. Results: After training, the detection network was evaluated on two datasets, resulting in classification accuracies of 97.3% and 98.3%, respectively. This network can identify Nocardia instances in about 24 milliseconds per image on a personal computer. The detection metrics of mAP50 on both datasets were 0.780 and 0.841, respectively. Conclusion: The Nocardia screening method can accurately and efficiently determine whether Nocardia exists in the images of Gram-stained sputum smears. Additionally, it can precisely locate the Nocardia instances, assisting doctors in confirming the presence of Nocardia.

Log (Lactobacillus crispatus/ Gardnerella vaginalis): a new indicator of diagnosing bacterial vaginosis.

To explore a new marker which can detect bacterial vaginosis (BV) with high sensitivity and specificity quantitatively. According to the Nugent Score, vaginal secretions from study participants were divided into BV, healthy, and BV-intermediate groups. First, we compared the obvious differences and high abundance of bacteria in the three groups using 16S rRNA-sequencing, and screened out candidate markers. Then, quantitative detection of these candidate markers from the three groups was done using real-time reverse transcription-quantitative polymerase chain reaction (RT-qPCR), followed by evaluation of the sensitivity and specificity. Finally, we verified the new markers using clinical cases. Gardnerella vaginalis, Atopobium vaginae, Lactobacillus, Megasphaera were screened out by 16S rRNA-sequencing. RT-qPCR data were transformed and analyzed through ROC curves. PCR results for these bacteria were log-transformed using Lactobacillus crispatus as the numerator and other BV-related bacteria as the denominator. Four new indicators were found. Of these, log L. crispatus/G. vaginalis (L/G) <0 was the best indicator. The sensitivity, specificity, positive predictive value, and negative predictive value of our system were 93.5%, 97.2%, 96.6 and 94.6%, respectively. Combination of data for 16S rRNA-sequencing and RT-qPCR revealed four indicators for BV detection. Of these, log L/G < 0 was the best indicator. Creating a molecular-diagnostic system independent of the Nugent Score for BV could have an important impact on the clinical management of BV.Abbreviation: log L. crispatus/G. vaginalis (logL/G); Bacterial vaginosis (BV); vaginal secretions (VSs); polymerase chain reaction (PCR); rRNA-sequencing (rRNA-seq); real-time reverse transcription-quantitative polymerase chain reaction (RT-qPCR); operational taxonomic unit (OTU); non-metric multidimensional scaling (NMDS); receiver operating characteristic (ROC).