Comparison of Proxy Indicators to Direct Measures Curated From Medical Records
Deep Learning Model Using NLP to Identify Metastatic Status from Unstructured Notes
Krishna Swaminathan, Principal AI Scientist, worked with other Concerto researchers and presents findings from the development of deep learning algorithms that can accurately impute metastatic status and site of metastasis from unstructured notes using NLP.
More Data Science & Machine Learning Abstracts
A Dynamic Model for Prediction of Metastatic Recurrence in Breast Cancer Patients
Machine Learning models that can dynamically predict risk of metastatic breast cancer (mBC) based on cumulative historical clinical data could help guide patient care and monitoring decisions. In this study, Vivek Vaidya, Senior Principal Scientist, shares results of analyses into an ML model that sought to predict risk of recurrence at any point in the patient journey.
A Machine Learning Model to Identify Lung Cancer Subtype
Smita Agrawal, PhD and Senior Director of Product Development, considers findings from a machine learning model to identify NSCLC patients from a heterogeneous cohort of lung cancer patients using Concerto's vast real-world database of structured Electronic Medical Records (EMR) data.
An AI Model to Impute ECOG Scores When Missing in Real-World Patient Data
Vivek Vaidya, Senior Principal Scientist, examines findings from a machine learning model built on structured data to impute ECOG scores using information from different points in the patient journey. ECOG Performance Scores are a strong prognostic indicator of outcomes. They are frequently unavailable in real-world settings but required for clinical trials.
Digital Patient Reported Outcomes Improve Quality of Care
Joanne Buzaglo, PhD and Executive Director of Concerto HealthAI's Patient Reported Outcomes (PRO) Solutions, discusses the effectiveness of using an electronic PRO system to facilitate compliance to required metric reporting and greater clinical efficiencies.