Data science use cases
Use Case 1: Deep Learning
Machine learning is a term that is often used but rarely understood. In this use case, wega has partnered up with an external collaborator to solve a concrete use case using machine learning: the automatic classification of Electrocardiograms (ECGs). In a clinical context, ECGs are a valuable tool for the diagnosis and prevention of cardiovascular diseases, which are the leading cause of death worldwide.
The collaborator is an internationally established provider of cardiac safety services. As such, they can provide the data as well as the clinical expertise that is necessary for training a deep neural network efficiently.
Combining this knowledge with the Data Science expertise at wega, we propose to create a deep neural network for the accurate, efficient and automatic interpretation of ECG signals; this model is termed ECG DNN.
The ECG DNN processes 12-lead ECG signals, and outputs a value between 0 and 1, which is used for classification.
Use Case 2: Cloud Data Management
Motivation
- Accessibility of clinical data is increasing rapidly, and companies must adapt to manage this data efficiently.
- Cloud solutions allow for easy storage of data , enable access to data across all company stakeholders in a controlled fashion.
- Any changes to data are recorded and readily available via an audit trail.
Proposal
- Data is collected from multiple sources. This collection of data is achieved via data pipelines in an automated fashion.
- The data undergoes several processing steps before being transferred to the Cloud Data Warehouse/Lake .
- Centralized storage enables collaboration between stakeholders through various data analytics use cases, all on a single platform.
wega - competencies
- wega can support you to implement the optimal data strategy, guidelines and quality standards.
- wega can guide you from data collection all the way to the final upload of clean data to the cloud.
- wega can bring you the most relevant Data Visualizations, Reports and Analyses to make more informed data-driven decisions.
Use case 3: Shiny Dashboard
This use case was created for an internal wega presentation to show the functionalities of an interactive R Shiny dashboard on a real dataset. Specifically, k-nearest neighbor and random forest algorithms were used to visualize the classification performance on this dataset in an interactive way.
The visualization can be accessed via the following link: Link.
Contact us
