A domain adaptation benchmark for T1-weighted brain magnetic resonance image segmentation.

Accurate brain segmentation is critical for magnetic resonance imaging (MRI) analysis pipelines. Machine-learning-based brain MR image segmentation methods are among the state-of-the-art techniques for this task. Nevertheless, the segmentations produced by machine learning models often degrade in the presence of expected domain shifts between the test and train sets data distributions. These domain shifts are expected due to several factors, such as scanner hardware and software differences, technology updates, and differences in MRI acquisition parameters. Domain adaptation (DA) methods can make machine learning models more resilient to these domain shifts. This paper proposes a benchmark for investigating DA techniques for brain MR image segmentation using data collected across sites with scanners from different vendors (Philips, Siemens, and General Electric). Our work provides labeled data, publicly available source code for a set of baseline and DA models, and a benchmark for assessing different brain MR image segmentation techniques. We applied the proposed benchmark to evaluate two segmentation tasks: skull-stripping; and white-matter, gray-matter, and cerebrospinal fluid segmentation, but the benchmark can be extended to other brain structures. Our main findings during the development of this benchmark are that there is not a single DA technique that consistently outperforms others, and hyperparameter tuning and computational times for these methods still pose a challenge before broader adoption of these methods in the clinical practice.

Analyzing Social Media to Explore the Attitudes and Behaviors Following the Announcement of Successful COVID-19 Vaccine Trials: Infodemiology Study.

BACKGROUND: The rollout of COVID-19 vaccines has brought vaccine hesitancy to the forefront in managing this pandemic. COVID-19 vaccine hesitancy is fundamentally different from that of other vaccines due to the new technologies being used, rapid development, and widespread global distribution. Attitudes on vaccines are largely driven by online information, particularly information on social media. The first step toward influencing attitudes about immunization is understanding the current patterns of communication that characterize the immunization debate on social media platforms. OBJECTIVE: We aimed to evaluate societal attitudes, communication trends, and barriers to COVID-19 vaccine uptake through social media content analysis to inform communication strategies promoting vaccine acceptance. METHODS: Social network analysis (SNA) and unsupervised machine learning were used to characterize COVID-19 vaccine content on Twitter globally. Tweets published in English and French were collected through the Twitter application programming interface between November 19 and 26, 2020, just following the announcement of initial COVID-19 vaccine trials. SNA was used to identify social media clusters expressing mistrustful opinions on COVID-19 vaccination. Based on the SNA results, an unsupervised machine learning approach to natural language processing using a sentence-level algorithm transfer function to detect semantic textual similarity was performed in order to identify the main themes of vaccine hesitancy. RESULTS: The tweets (n=636,516) identified that the main themes driving the vaccine hesitancy conversation were concerns of safety, efficacy, and freedom, and mistrust in institutions (either the government or multinational corporations). A main theme was the safety and efficacy of mRNA technology and side effects. The conversation around efficacy was that vaccines were unlikely to completely rid the population of COVID-19, polymerase chain reaction testing is flawed, and there is no indication of long-term T-cell immunity for COVID-19. Nearly one-third (45,628/146,191, 31.2%) of the conversations on COVID-19 vaccine hesitancy clusters expressed concerns for freedom or mistrust of institutions (either the government or multinational corporations) and nearly a quarter (34,756/146,191, 23.8%) expressed criticism toward the government's handling of the pandemic. CONCLUSIONS: Social media content analysis combined with social network analysis provides insights into the themes of the vaccination conversation on Twitter. The themes of safety, efficacy, and trust in institutions will need to be considered, as targeted outreach programs and intervention strategies are deployed on Twitter to improve the uptake of COVID-19 vaccination.