Matching patients to their medical records from multiple health care providers is critical to medical care, but can be challenging to accomplish because their records can be incomplete or inaccurate, and patients often share similar names. How, for example, to match medical records to the correct “John Jones” or “Maria Garcia” from their primary care doctor's office, the lab which processed tests the doctor ordered, the imaging center where they had a cancer screening, the out of town hospital where they were treated while on vacation? What if a name is recorded as James at one site and as Jim at another? And what if a common or uncommon name is mistyped at one or more places?...
Huiping Xu
See the following -
Open Source Machine Learning Tools are as Good as Humans' in Cancer Surveillance According to Regenstrief, Indiana Univ. Study
Press Release |
Indiana University |
April 21, 2016
Machine learning has come of age in public health reporting according to researchers from the Regenstrief Institute and Indiana University School of Informatics and Computing at Indiana University-Purdue University Indianapolis. They have found that existing algorithms and open source machine learning tools were as good as, or better than, human reviewers in detecting cancer cases using data from free-text pathology reports. The computerized approach was also faster and less resource intensive in comparison to human counterparts.
- Login to post comments