Publications by authors named "B Mortazavi"
, particularly uncultured representatives, are one of the most abundant microbial groups in coastal salt marshes, dominating the belowground rhizosphere, where over half of plant biomass production occurs. However, this class generally remains poorly understood, particularly in a salt marsh context. Here, novel metagenome-assembled genomes (MAGs) were generated from the salt marsh rhizosphere representing , , JAAYZQ01, B4-G1, JAFGEY01, UCB3, and orders.
View Article and Find Full Text PDFArticle Synopsis
- Leptomeningeal carcinomatosis is a rare metastatic pattern in genitourinary cancer, found in less than 0.1% of cases, and can occur even after initial treatments with enfortumab vedotin (EV).
- Two cases of metastatic urothelial cancer are presented: both patients initially showed positive responses to EV but later developed severe neurologic symptoms due to leptomeningeal metastases confirmed through imaging and cytology.
- The cases highlight an unusual progression pattern among patients treated with EV, suggesting the need for further investigation into this type of cancer spread.
Article Synopsis
- Hypertension is a major risk factor for serious health conditions, and there’s potential for artificial intelligence (AI) to improve how it's diagnosed and managed.* -
- AI technologies, particularly machine learning, could personalize treatment and enhance blood pressure monitoring, but effective collaboration among health professionals and data scientists is crucial.* -
- A workshop by the National Heart, Lung, and Blood Institute highlighted communication gaps in healthcare, innovative methods for managing hypertension, and challenges to implementing AI in real-world settings.*
Article Synopsis
- Advances in self-supervised learning (SSL) have improved medical image diagnosis using limited labeled datasets, but current methods mainly focus on still images, not video modalities like echocardiography.
- The EchoCLR approach was developed to enhance echocardiogram video analysis through techniques like contrastive learning and frame reordering, aiming for better performance in diagnosing cardiac diseases.
- Results showed that models pretrained with EchoCLR significantly outperformed standard transfer learning methods in classifying left ventricular hypertrophy and aortic stenosis, demonstrating SSL's effectiveness in achieving high accuracy with minimal labeled data.*
To establish Pulse2AI as a reproducible data preprocessing framework for pulsatile signals that generate high-quality machine-learning-ready datasets from raw wearable recordings. We proposed an end-to-end data preprocessing framework that adapts multiple pulsatile signal modalities and generates machine-learning-ready datasets agnostic to downstream medical tasks. a dataset preprocessed by Pulse2AI improved systolic blood pressure estimation by 29.
View Article and Find Full Text PDF