MAP - Multimodal Automated Phenotyping

Electronic health records (EHR) linked with biorepositories are a powerful platform for translational studies. A major bottleneck exists in the ability to phenotype patients accurately and efficiently. Towards that end, we developed an automated high-throughput phenotyping method integrating International Classification of Diseases (ICD) codes and narrative data extracted using natural language processing (NLP). Specifically, our proposed method, called MAP (Map Automated Phenotyping algorithm), fits an ensemble of latent mixture models on aggregated ICD and NLP counts along with healthcare utilization. The MAP algorithm yields a predicted probability of phenotype for each patient and a threshold for classifying subjects with phenotype yes/no (See Katherine P. Liao, et al. (2019) <doi:10.1093/jamia/ocz066>.).

Last updated 22 days ago

8.58 score 6 stars 1 dependents 177 scripts 519 downloads

PheCAP - High-Throughput Phenotyping with EHR using a Common Automated Pipeline

Implement surrogate-assisted feature extraction (SAFE) and common machine learning approaches to train and validate phenotyping models. Background and details about the methods can be found at Zhang et al. (2019) <doi:10.1038/s41596-019-0227-6>, Yu et al. (2017) <doi:10.1093/jamia/ocw135>, and Liao et al. (2015) <doi:10.1136/bmj.h1885>.

Last updated 4 years ago

6.02 score 21 stars 8 scripts 140 downloads

phecodemap - Visualization for Phecode Mapping with ICD-9 and ICD-10-cm Codes

Phecodemap builds a shiny app to visualize the hierarchy of Phecode Mapping with ICD. The same Phecode hierarchy is displayed in two ways: as a sunburst plot and as a tree.

Last updated 11 months ago

5.55 score 3 stars 13 scripts 142 downloads

PheNorm - Unsupervised Gold-Standard Label Free Phenotyping Algorithm for EHR Data

The algorithm combines the most predictive variable, such as count of the main International Classification of Diseases (ICD) codes, and other Electronic Health Record (EHR) features (e.g. health utilization and processed clinical note data), to obtain a score for accurate risk prediction and disease classification. In particular, it normalizes the surrogate to resemble gaussian mixture and leverages the remaining features through random corruption denoising. Background and details about the method can be found at Yu et al. (2018) <doi:10.1093/jamia/ocx111>.

Last updated 4 years ago

4.70 score 5 stars 7 scripts 203 downloads

SCORNET - Semi-Supervised Calibration of Risk with Noisy Event Times

A consistent, semi-supervised, non-parametric survival curve estimator optimized for efficient use of Electronic Health Record (EHR) data with a limited number of current status labels. See van der Laan and Robins (1997) <doi:10.2307/2670119>.

Last updated 4 years ago

cpp

4.48 score 3 stars 3 scripts 112 downloads

SAMGEP - A Semi-Supervised Method for Prediction of Phenotype Event Times

A novel semi-supervised machine learning algorithm to predict phenotype event times using Electronic Health Record (EHR) data.

Last updated 4 years ago

openblascpp

4.30 score 2 stars 2 scripts 134 downloads

kesernetwork - Visualization of the KESER Network

A shiny app to visualize the knowledge networks for the code concepts. Using co-occurrence matrices of EHR codes from Veterans Affairs (VA) and Massachusetts General Brigham (MGB), the knowledge extraction via sparse embedding regression (KESER) algorithm was used to construct knowledge networks for the code concepts. Background and details about the method can be found at Chuan et al. (2021) <doi:10.1038/s41746-021-00519-z>.

Last updated 2 years ago

4.00 score 1 stars 7 scripts 222 downloads

sureLDA - A Novel Multi-Disease Automated Phenotyping Method for the Electronic Health Record

A Novel Multi-Disease Automated Phenotyping Method for the Electronic Health Record.

Last updated 4 years ago

cpp

3.60 score 4 stars 3 scripts 133 downloads