Mission and Goals
Fetal ECG Extraction
Electronic Nose Sensing
(AI, ANN, DSP)
Technical (Software/ Hardware)
1. Hardware Expertise:
- Packages and Technologies
Even with the popular use of software to provide solutions for many of today’s problems, we feel that an engineer is not totally complete without also having an understanding of hardware. Our Convergent Engineering team has researchers and engineers who are extremely proficient in designing, building
and testing highly complex electrical and electronic hardware systems. Two examples of ND's design and prototyping experience are our 'Fetal ECG Amplifier' and real-time 'Work Of Breathing System'.
Fetal ECG Amplifier:
Recording of fetal ECG signals from electrodes placed on the maternal abdomen is a very challenging problem.
The fetal ECG is many orders of magnitude smaller than the maternal ECG and is typically embedded in noise. The ECG recorded from the maternal
abdomen is coupled with both maternal and fetal ECG. The problem boils down to amplifying the raw ECG signal to maximize the signal to noise
ratio (SNR) of the fetal signal, while at the same time preventing over-amplification of maternal ECG and saturation of ECG amplifier. At
Convergent, our engineering team has developed a high resolution, low noise ECG amplifier tailored for
collecting ultra low amplitude signals
such as fetal ECG. The amplifier is an 8-channel unipolar design where all eight signals are measured with respect to a reference electrode. The
amplifier also uses driven right leg (DRL) circuitry to reduce common mode noise between the patient and the amplifier common. The amplifier has a
bandwidth between 0.1 Hz and 100 Hz, with a 60 Hz notch. The gain of the amplifier is variable between 2000 and 8000. The amplifier runs from a wall
power supply, but has onboard isolations circuits that minimize leakage currents and protect the subject from harm in case of an electrical short.
Typically the amplifier output is interfaced to a 16-bit analog to digital converter (ADC) that samples the data at 200 Hz.
We are currently developing the next generation of fetal ECG amplifier. Unlike previous designs that relied mainly on amplifier
front-end and analog filters for recording fetal ECG, the new design will be a hybrid system utilizing both analog hardware circuitry and digital signal processing.
A simplified ECG amplifier front-end will interface to the electrodes connected to the patient and transmit raw ECG signals from to a very high resolution ADC.
The digitized data will be transferred to a DSP, which will process the data through 60 Hz notch and 100 Hz low pass digital filters.
Work of Breathing (WOB) Hardware:
(Ventilator Project HW)
The WOB hardware unit measures the pressure changes from esophageal and tracheal catheters in patients receiving mechanical ventilation.
The analog pressure signals from the catheters are digitized and processed before transmission to host a computer via the RS232 serial interface.
Control of the WOB hardware is also done from the host computer through the serial interface. Additionally, the user also has the flexibility to develop
custom applications in Visual Basic to enable communication and control of the WOB hardware. Pneumatic valves and algorithms are included to automatically zero and purge the catheters when necessary.
Additionally, the WOB hardware can also interface to the Hewlett Packard (HP) Component Monitoring System (CMS), also known as the Merlin, to record any measurable
signal on the Merlin. The WOB hardware can read two signals from the Merlin via the analog interface connector and transmit the data to the host PC.
(Back To Top)
2. Software Expertise:
Developing a neural network model requires large amount of data for training, testing and validation. Thus, the ‘data’ aspect is an extremely critical component in all of our research projects. This data can be any entity that is valuable for the analysis and outcome - for example, the Height of a Patient or a Ventilator Pressure Support Setting or a Continuous Signal from ECG Electrodes. The numbers or signals by themselves generally don't make much sense. Also, real-world signals are too noisy to be analyzed directly. Hence, there are many steps involved in the entire process of understanding the data and we have developed tools to automate each of them.
The typical process looks like this:
Convergent Engineering has developed data collection software tools for various projects. Although the objective of each project was different, the main components of these tools remained the same.
Interfacing with Medical Devices: It is very important to understand the medical device from which you are collecting data and determine what exactly you need. We have developed software programs that interface with wide range of medical instruments - from getting values for only a few parameters from a Pulse-Oximeter to getting thousands of them from a NICO every second.
Signal processing: Most of the time real world data is too noisy to be useful. Hence several filters are added directly into the data collection software or hardware to filter out noise and store cleaner data.
Data storage: It is vital to store data in a structured fashion so that it can be presented/analyzed in intuitive way and with ease. Thus, we have developed a sophisticated methodology for designing and developing structured databases to store data.
Real-time data presentation: It's always important to view data in a real-time fashion in order to receive immediate feedback. Specially, in clinical research projects it is vital to know the consequences of a change immediately. Our tools show incoming Data in charts and numbers in a real-time fashion while collecting and storing new data every 200th of a second (0.005 Sec).
(Vent Project Data Collection)
(Pulse-Ox Data Collection)
Convergent Engineering meets FDA (21) CFR Part 11 recommendations and HIPAA (Health Insurance Portability and Accountability Act) requirements and considers patient privacy, data accuracy and security to be top priorities. Since data collection is carried out at various different physical locations, we have also developed tools to transfer secured data over the Internet.
As stated earlier, real-world signals have a significant amount of noise in them. Before we analyze the data we need to filter out that noise. Also we may need to do some calculations which are unnecesssary while collecting data. NThus, we have developed a set of post-processing tools which work on raw data to produce clean, consistent, informative and structured data.
(ECG Project Viewer)
(Vent Project Viewer)
A Viewer is a tool which lets you view the data offline. Real-time data representation is necessary but is not efficient when you have to look at the data more closely. Our offline viewer tools allow users to move to any section of data, view data, play data in forward and reverse motion at different speeds, take and print snapshots,
do calculations while viewing etc. The viewer also interfaces with popular worksheet and presentation software to put data/charts/snapshots directly in a workbook or slide.
The Analyzer is a tool that compliments a viewer. When an expert is viewing data using the viewer, the analyzer will help him take notes about the sections of data.
The analyzer will let the expert find good/bad regions in the data, store the regions, store the notes, calculate a variety of parameters etc. It lets you
compare different sections or different parameters. It keeps the information in the database which will be retrieved in the next session. The combination of
Analyzer with viewer is critical for extracting information from the data and making sense of it all.
(Analyzing Valid Data)
Convergent Engineering has experience in developing reliable and useful software tools for a variety of the steps in a research project. If you are interested in a customized data collection system, please contact Convergent Engineering at firstname.lastname@example.org.
(Back To Top)
3. Packages and Technologies:
The Convergent Engineering team is proficient in working with several stand-alone commercial software packages. It also has experience in developing tools to facilitate interaction between the packages. For example, some of the tools our team is proficient in include NeuroSolutions, MATLAB, Microsoft Office, Microsoft Visual Studio, Component One Studio, etc. We use various software technologies such as Internet Data Transfer, Databases and Wireless Communication. Convergent Engineering has been successful in collecting data from medical devices without cables using Bluetooth technology to simplify the process of setting up data collection stations. We also have experience in programming for PDAs, Tablet PCs and Smart Phones.
(Back To Top)
ConvEng Home |
Current Projects |
Questions regarding Projects, Applications, or Partnerships? Contact email@example.com
Web Site Design and Implementation Copyright © 2007 Convergent Engineering, Inc.