Publications by authors named "T DeSilvio"
Batch effects (BEs) refer to systematic technical differences in data collection unrelated to biological variations whose noise is shown to negatively impact machine learning (ML) model generalizability. Here we release CohortFinder (http://cohortfinder.com), an open-source tool aimed at mitigating BEs via data-driven cohort partitioning.
View Article and Find Full Text PDFArticle Synopsis
- The study focuses on improving the evaluation of locally advanced rectal cancers by using U-Net deep learning models to automatically segment critical rectal structures from MRI scans, aiming to ease the labor-intensive manual annotation process.
- Results showed that the region-specific U-Net models achieved high accuracy in segmenting the outer rectal wall and lumen, performing similarly to multiple trained readers, and outperforming a standard multi-class U-Net model by an average of 20%.
- This advancement in deep learning for MRI analysis can lead to more precise tumor evaluations and enhance the development of image-based tools essential for treating rectal cancers after neoadjuvant therapy.
Rich sources of obesity-related data arising from sensors, smartphone apps, electronic medical health records and insurance data can bring new insights for understanding, preventing and treating obesity. For such large datasets, machine learning provides sophisticated and elegant tools to describe, classify and predict obesity-related risks and outcomes. Here, we review machine learning methods that predict and/or classify such as linear and logistic regression, artificial neural networks, deep learning and decision tree analysis.
View Article and Find Full Text PDF