Single vs. dual head using geometric mean and its effect on gastric emptying time.

Objectives Gastric emptying (GE) scintigraphy demonstrates the rate in which ingested foods and/or liquids leave the stomach and enter the duodenum of the small intestine. While there are other modalities that can be used for this study, nuclear scintigraphy manifests the most qualitative results. Because the stomach is located both anteriorly and posteriorly in the human body, it is ideal to acquire projections using both views. Several departments have been utilizing single detectors for efficiency purposes, which underestimates and overestimates GE times. This research analyzes the results of ten GE studies by calculating geometric mean along with a decay factor, and comparing anterior versus anterior and posterior acquisitions.

Methods 0.500 µCi (18,500 Bq) of Tc-99m sulfur colloid was injected into the middle of an egg salad sandwich and consumed by the patient. A dynamic imaging study was obtained for 90 frames at 1 minute per frame. Head 1 of the detector was used to acquire anterior images, while head 2 was used to acquire posterior images. A total of ten patients’ GE studies were analyzed via hand-drawn regions of interest (ROIs) in both the anterior view and anterior/posterior views. Mean GE time was compared between the two methods of processing to determine if the results were significant enough to suggest a new diagnosis.

Results One out of ten patients’ results were changed from abnormal to normal. Symptoms of gastroparesis and dyspepsia occurred in females slightly more than in males.

Conclusions Dual headed detectors, which factor in a geometric mean, were determined to be the most accurate means of conducting a gastric emptying study. A limitation to the study was that while processing the data, the ROIs could not be copied from the anterior/posterior acquisition and pasted to the anterior acquisition. This meant that a new ROI had to be drawn for each patient’s data which was not the most accurate or consistent method. Another limitation was having a small sample size.