Peter R Mouton

Title: Peter R. Mouton and Automated Stereology

Introduction

Peter R. Mouton is an inventor based in Tampa, Florida. His patent work addresses automated stereology, a method for analyzing tissue characteristics using image data.

Latest Patents

Mouton is associated with 1 patent, titled “Automated Stereology for Determining Tissue Characteristics.” The invention describes systems and methods that use deep learning to make stereology analysis more efficient and accurate. A method may capture a stack of images of a three-dimensional object, assess whether an existing model is suitable, preprocess the data, and train or retune a deep learning model. The model then assigns confidence scores to labels, which a user can review before selected labels are added to the training set. The trained model is subsequently used to perform stereology analysis.

Career Highlights

The patent highlights an approach to improving deep learning models for tissue analysis. It describes active learning to support more efficient model training, as well as retuning existing models to improve accuracy for a particular data set.

Collaborations

Mouton’s listed coworkers include Saeed S. Alahmari and Lawrence O. Hall.

Conclusion

Peter R. Mouton’s patent focuses on applying deep learning to automate and refine stereology analysis of tissue images.