Content Analysis and Quality Evaluation of Cesarean Delivery-Related Videos on YouTube: Cross-sectional Study.

BACKGROUND: YouTube is one of the most popular open-access video-sharing websites, and it is also used to obtain health care information. Cesarean delivery is the most common major surgical intervention in many countries. Videos related to cesarean delivery have also been uploaded to YouTube. However, no study has explored the overall quality of cesarean delivery videos on the platform. OBJECTIVE: The objective of this study was to analyze the content and evaluate the quality of the most frequently viewed videos related to cesarean delivery that are accessible on YouTube. METHODS: We searched for a total of 18 terms by combining the 6 terms retrieved from Google AdWords and the 3 terms c section, cesarean section, and cesarean delivery, which are used interchangeably. Videos were sorted by view count, and the 100 videos with the highest view counts were chosen. The number of views, duration, likes and dislikes, content type, and source of each video were recorded. In evaluating the quality of the videos, we referred to a previous study. Additionally, we developed a detailed scoring method that comprehensively evaluates the videos related to cesarean delivery by including the necessary information for each element of the cesarean delivery and whether scientific evidence was presented. RESULTS: Of the 100 videos analyzed, the most prevalent content (n=28) was videos that contained the actual surgical procedure of a cesarean delivery, and the most common source of cesarean delivery videos was physicians (n=30). Videos directly related to cesarean delivery, such as explanation of the surgery and the actual surgical procedure, were mainly uploaded by medical groups and scored higher than the videos indirectly related to cesarean delivery, which were mainly uploaded by nonmedical groups. In addition, videos directly related to cesarean delivery were more often uploaded earlier in time, with lower like ratios compared to indirect videos. CONCLUSIONS: YouTube is currently not an appropriate source for patients seeking information on cesarean delivery.

Classification of cervical neoplasms on colposcopic photography using deep learning.

Colposcopy is widely used to detect cervical cancers, but experienced physicians who are needed for an accurate diagnosis are lacking in developing countries. Artificial intelligence (AI) has been recently used in computer-aided diagnosis showing remarkable promise. In this study, we developed and validated deep learning models to automatically classify cervical neoplasms on colposcopic photographs. Pre-trained convolutional neural networks were fine-tuned for two grading systems: the cervical intraepithelial neoplasia (CIN) system and the lower anogenital squamous terminology (LAST) system. The multi-class classification accuracies of the networks for the CIN system in the test dataset were 48.6 ± 1.3% by Inception-Resnet-v2 and 51.7 ± 5.2% by Resnet-152. The accuracies for the LAST system were 71.8 ± 1.8% and 74.7 ± 1.8%, respectively. The area under the curve (AUC) for discriminating high-risk lesions from low-risk lesions by Resnet-152 was 0.781 ± 0.020 for the CIN system and 0.708 ± 0.024 for the LAST system. The lesions requiring biopsy were also detected efficiently (AUC, 0.947 ± 0.030 by Resnet-152), and presented meaningfully on attention maps. These results may indicate the potential of the application of AI for automated reading of colposcopic photographs.

A Comparison between the Anterior and Posterior Approach to US-guided Shoulder Articular Injections for MR Arthrography

PURPOSE: To assess the feasibility of ultrasound-guided shoulder joint injections by the anterior and posterior approaches for performing a MR arthrography. MATERIALS AND METHODS: Between April 2007 and June 2008, 28 patients underwent ultrasound-guided intra-articular contrast injections for a MR arthrography. This study was approached using the antegrade method. As well, all injections were performed by one radiologist. The patient selection criteria for the anterior and posterior approaches were randomly for the anterior approach and alternatively for the posterior approach. Each approach included 13 cases and each were injected by the anterior or posterior approaches exclusively . The patients were classified into three groups as follows: 1) the successful intra-articular injection group, 2) the small leakage group, and 3) the injection failure group. RESULTS: Of the 15 patients that underwent the anterior approach, two were unsuccessful for the MR arthrography. Whereas, two other cases experienced small leakage. The two failed MR arthography cases using the anterior approach were later injected using the posterior approach. The 15 patients subjected to the posterior approach to perform a MR arthrogram were successful in all cases, without any leakage. CONCLUSION: The ultrasound-guided intra-articular injections were feasible with a high success rate by both the anterior and posterior approaches. However, better results were achieved from the posterior approach than the anterior approach in this study.