Publications by authors named "P Clopton"
Background: Large language models (LLMs), such as ChatGPT, excel at interpreting unstructured data from public sources, yet are limited when responding to queries on private repositories, such as electronic health records (EHRs). We hypothesized that prompt engineering could enhance the accuracy of LLMs for interpreting EHR data without requiring domain knowledge, thus expanding their utility for patients and personalized diagnostics.
Methods: We designed and systematically tested prompt engineering techniques to improve the ability of LLMs to interpret EHRs for nuanced diagnostic questions, referenced to a panel of medical experts.
Article Synopsis
- Segmenting CT is essential for clinical practices like personalized cardiac ablation, but traditional machine learning methods often require large labeled datasets which are difficult to gather.
- The article introduces the DOKEN algorithm, which uses domain knowledge to automatically label a small training set, enabling high-performance ML segmentation without the need for extensive data.
- In tests, the DOKEN-enhanced nnU-Net model showed impressive segmentation results, achieving a high Dice score of 96.7% and demonstrating performance comparable to expert manual segmentation, thus validating its efficacy in real-world applications.
Women Trainees in Electrophysiology and the Effect of Role Models.
Circ Arrhythm Electrophysiol
June 2024
Background: Risk stratification for ventricular arrhythmias currently relies on static measurements that fail to adequately capture dynamic interactions between arrhythmic substrate and triggers over time. We trained and internally validated a dynamic machine learning (ML) model and neural network that extracted features from longitudinally collected electrocardiograms (ECG), and used these to predict the risk of malignant ventricular arrhythmias.
Methods: A multicentre study in patients implanted with an implantable cardioverter-defibrillator (ICD) between 2007 and 2021 in two academic hospitals was performed.
Article Synopsis
- Segmentation of cardiac CT is crucial for procedures like cardiac ablation, but traditional machine learning methods require extensive labeled data, which is hard to gather.
- The researchers developed a "virtual dissection" model that uses simple geometric shapes to represent atrial anatomy, enabling effective segmentation with minimal training data in a sample of just 6 digital hearts.
- Their results showed high accuracy in segmenting atrial structures across multiple datasets, significantly reducing segmentation time in live patients while maintaining accuracy comparable to expert assessments.