Melax | Products

For NLP Development and Deployment:

● Model Training- DeepMed

DeepMed Docker

Our deep learning solutions - DeepMed provides a docker container that allows for an easy-to-use training of deep-learning models with a library of state-of-the-art algorithms(e.g., the BERT models) and automated parameter optimization.

DeepMed for SaaS

DeepMed is also available as a Software as a Service(SaaS) platform for Cloud-based computing with integrated state-of-the-art deep learning algorithms and improved Interoperability. It greatly shortens the time and reduces cost for clinical NLP development.

● Hybrid Pipeline - CLAMP GUI

The CLAMP GUI(Graphical User Interface) system provides not only high-performance NLP modules and pipelines customized for clinical text, but also user-friendly interfaces for quickly building customized NLP pipelines. It provides multiple useful features for NLP professionals.
- Annotate local data and re-train machine learning models with a few clicks
- Specify rules on the top of ML models to improve performance
- Export NLP pipelines as web services for easy integration
- Provide Java APIs for individual components for system integration

● NLP on Cloud - CLAMP CS

The CLAMP CMD(Command Line) is an off-the-shelf general-purpose clinical NLP system built on proven methods with good performance and high speed for extracting general clinical concepts such as diseases, drugs, labs, and procedures.

CLAMP components are built on proven methods in many clinical NLP challenges
Identify context of entities such as negation, certainty, and values
Support Multi-threads/Multi-processes/Docker/Distributed systems
Process local files/Database tables/HDFS files

For NLP Applications:

● EHR Chart Review - Vital

By leveraging state-of-the-art information retrieval and information extraction technologies, Vital provides a powerful yet user-friendly platform for quickly searching millions of clinical documents in EHRs, thus to efficiently identify cohorts of patients of interest, as well as specific phenotypes for each patient in a cohort.

Use cases of Vital include healthcare quality reporting, disease registry curation, and clinical and translational research. Our studies have shown that Vital dramatically improves efficiency of manual chart review in both operational and research tasks.

Schedule processing tasks
Allocate computational resources for each task
Search and visualize outputs of NLP outputs
Support Common Data Models (CDM) such as OMOP CDM

● Literature Systematic Review

A systematic review of literature uses an objective and transparent approach for research synthesis to minimize bias, by providing a complete, exhaustive summary of current literature relevant to a research question. The literature systematic review tool follows a typical workflow:

Create a structured question (and queries) to guide the review
Perform a thorough search of the literature for relevant papers automatically
Display databases and citation indexes such as Web of Science, Embase, and PubMed and any individual journals that were searched.
Create eligibility criteria: inclusion and exclusion
Using deep learning based methods to re-rank the searched literature based on the eligibility criteria and relevance
Assign an objective assessment of methodological quality preferably by using methods conforming to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (the current guideline) or the high quality standards of Cochrane

● Melax Marketplace

The Melax Knowledge Sharing Marketplace allows users to share pipelines developed by themselves or to purchase pipelines of their interest.
These pipelines include Cancer pipeline, Substance Overdose pipeline, Social Determinants pipeline, and many more...