RESUMO
Artificial intelligence (AI) and machine learning (ML) show promise in various medical domains, including medical imaging, precise diagnoses, and pharmaceutical research. In neuroscience and neurosurgery, AI/ML advancements enhance brain-computer interfaces, neuroprosthetics, and surgical planning. They are poised to revolutionize neuroregeneration by unraveling the nervous system's complexities. However, research on AI/ML in neuroregeneration is fragmented, necessitating a comprehensive review. Adhering to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) recommendations, 19 English-language papers focusing on AI/ML in neuroregeneration were selected from a total of 247. Two researchers independently conducted data extraction and quality assessment using the Mixed Methods Appraisal Tool (MMAT) 2018. Eight studies were deemed high quality, 10 moderate, and four low. Primary goals included diagnosing neurological disorders (35%), robotic rehabilitation (18%), and drug discovery (12% each). Methods ranged from analyzing imaging data (24%) to animal models (24%) and electronic health records (12%). Deep learning accounted for 41% of AI/ML techniques, while standard ML algorithms constituted 29%. The review underscores the growing interest in AI/ML for neuroregenerative medicine, with increasing publications. These technologies aid in diagnosing diseases and facilitating functional recovery through robotics and targeted stimulation. AI-driven drug discovery holds promise for identifying neuroregenerative therapies. Nonetheless, addressing existing limitations remains crucial in this rapidly evolving field.
RESUMO
Background and objective Stroke-related deaths have been one of the major causes of death worldwide due to its rising risk factors. As a result, several people rely on YouTube as a great source of information on stroke without knowing the genuineness of the content. This study aims to assess the quality and reliability of the information on stroke uploaded on the YouTube platform using the Global Quality score (GQS) and DISCERN score (DS), respectively. Methodology A cross-sectional observational study was conducted in April 2023. Stroke-related keywords were used to search for videos on YouTube. Videos that met inclusion criteria were evaluated for baseline characteristics (likes, comments, views, duration of video, time since posted, and uploader type) and type of information in the video about stroke (symptoms, etiology, treatment, and other parameters). These videos were then evaluated for quality and reliability of information using GQS and DS, respectively. Results After applying inclusion and exclusion criteria and removing the duplicates, 73 YouTube videos were selected. The videos had a total number of 23,927,445 views, 385,324 likes, and 31,927 comments. Maximum videos were uploaded by hospitals (25, 34.2%). Several videos described the symptoms (54, 73.97%), treatment (50, 68.49%), and etiology (49, 67.12%) of stroke. The reach of videos measured by the Video Power Index (VPI) was highest for videos uploaded by healthcare organizations (mean VPI = 120.11). There was no statistically significant difference (P > 0.05) in the quality (GQS score) and reliability (DS) of videos uploaded by doctors, hospitals, healthcare organizations, and news channels. Conclusions YouTube can become an important source to disseminate information about health-related conditions like stroke if the videos uploaded are of high quality (GQS score) and reliable (DS).
