Data mining techniques simplify PET report evaluations

Thursday, December 2 | 12:45 p.m.-1:15 p.m. | LL-INE1153-THB | Lakeside Learning Center
In this poster presentation, data mining techniques to classify the complexity and time of review of PET scan reports will be explained.

Reading and dictating a PET scan report can be a time-consuming procedure, which can take longer than the relative value units (RVUs) assigned to it to calculate financial reimbursement. Therefore, Murthy Chamarthy, MD, a radiologist at Montefiore Medical Center in New York City, and colleagues conducted a study using data mining techniques and algorithms to assess the time taken to review a PET scan and the complexity of the reports.

Chamarthy will discuss the study undertaken between November 2009 and March 2010 that analyzed 428 reports, and explain how various data mining algorithms were used to classify the data. Almost half of the reports took more than 20 minutes to interpret, and 16% were classified as being complex.

The researchers plan to validate the 34 keywords they identified and the techniques they used in a future, much larger study, and Chamarthy will also comment on these plans.

Page 1 of 603
Next Page