California Scientific  *  BrainMaker Neural Network Software  *   800-284-8112 * 916-372-6800
California Scientific      BrainMaker Neural Network Software
Home Applications What Are Neural Networks Products Ordering Info Tech Support
Applications

Stocks
Currency Predictions
Corporate Bonds
S&P 500 Index
Stock Price Prediction
Stock Prophet
Natural Gas Prices

Business
Direct Mail
Credit Scoring
Real Estate Appraisal
Bad Cops
Finding Gold
Jury Summoning
Highway Maintenance

Medical
Breast Cancer
Hospital Expenses
Functional Recovery
Heart Attack Diagnosis
ER Test ordering
Psychiactric Care
Giant Cell Arteritis

Sports
Thoroughbreds
Thoroughbreds 1
Dog Racing

Science
Solar Flares
Protein Sequencing
Mosquito Identification
Spectroscopy
Weather Forecasting
Predicting El Nino
Air Quality

Manufacturing
Welding Quality
Plastics Testing
IC Quality Control
Paper Quality
Concrete Testing
Beer Quality
Enzyme Synthesis
Steam Quality

Pattern Recog.
Speech Recognition
Article Classification
Working with Chaos
Chemical Drawings
Sequence Prediction

BrainMaker Orders Medical Laboratory Tests for ER

When a patient appears at the emergency room door it is sometimes an hour until a doctor can see him or her. It may be another hour until the lab can do the ordered tests. In order to save patient waiting time, Dr. Steven Berkov of Kaiser Hospital in Walnut Creek, California, developed a neural network program that can order the lab tests as soon as the patient is admitted. Up to 38 lab tests can be ordered by the neural network.

Not only does the neural network save up to two hours of patient waiting time, it can reduce expenses. When the pilot system was tested, it reduced the number of tests that were ordered by 10-15%. Dr. Berkov says it could save half a million dollars a year. The neural network is able to reduce the number of tests for two reasons. First, medical records are used as examples for training the neural network. It can be determined which tests were actually necessary in retrospect, so the neural network can be trained to order only the pertinent tests. Second, nurses had been given the blanket permission to order tests and they tended to order even more than doctors.

The neural network has 67 inputs that include patient demographics and symptoms. This information is gathered when the patient is admitted and placed in the medical record. There are 38 outputs, each representing a different test that might be ordered. The pilot system neural network was trained on 250 patients from past hospital medical records.

When the pilot system was tested on new patients it was found to be about 95% accurate, according to Dr. Berkov. Most of the time the neural network would order most of the necessary tests. Sometimes it did not order enough, but Dr. Berkov explained that usually the doctor would only need to call the lab and order another test on the already collected specimen.

Dr. Berkov is working on a neural network that is even more robust and focused. He is waiting for funds from the hospital to pursue this project on a larger scale and implement it in the hospital.

INPUTS
age
sex
critical
serious
routine
ambulance
blood pressure
temperature
medications
diarrhea
lower abdominal pain
upper abdominal pain
arrhythmia
chills
GI bleed
vomiting
headache
infection
etc.

OUTPUTS
blood gasses
acetone
Arteritis (ANA)
CO2
EKG
glucose
potassium
sodium
liver (SGOT)
liver (SGPT)
stool
type & cross blood
etc.

Home Applications What Are Neural Networks Products Ordering Info Tech Support
California Scientific      BrainMaker Neural Network Software

California Scientific
Sales@CalSci.com • 800-284-8112 • 916-372-6800