Reverse Engineering Ultrasound Files to Model Pelvic Organs

INTRODUCTION Pelvic floor dysfunction affects 25% of women within the United States [1]. Pelvic floor dysfunction encompasses a group of disorders that occur because of weakened pelvic muscles and fascia after damage or trauma. Most significantly, a compromised levator ani muscle is thought to be responsible for the onset of these pelvic floor problems. By the year 2050, the number of women with at least one pelvic floor disorder is predicted to grow from 28.1 million to 43.8 million [1]. Sensation, pain, discomfort, and organ movement all contribute to the severity of these disorders. For these reasons, pelvic floor dysfunction has profound impacts on quality of life. Imaging methods give valuable insight into the functionality and structural interdependence of pelvic organs [2]. Imaging is the best way to successfully diagnose and detection pelvic floor dysfunction because effectively studying imaging data with the proper tools of assessment compensates for the failures made during human evaluations [3]. Magnetic resonance imaging (MRI) and ultrasound sonography are commonly used imaging techniques for observing problems within the pelvic floor. Because MRI tends to be more costly and less regularly used for routine assessment while ultrasound is done at the point of care by providing dynamic images, ultrasound is a viable modality through which to study pelvic floor disorders. As ultrasound imaging is increasingly being regarded as a safer, less expensive, and more patient-friendly, it plays an important role in studying and diagnosing pelvic muscle and fascia deterioration [2]. Inside the Tissue Mechanics Laboratory, computational models, made from 3D endovaginal diagnostic images, that replicate real patient anatomy are often created and used for the purpose of evaluating complex mechanical systems. Figure 1 provides a high-level overview of the process of creating a computational model from imaged anatomy. This method of observing pathophysiology gives better understanding to the functional and structural interactions that affect the diagnostic and treatment tools. Sometimes, after obtaining ultrasound records and using them to study pelvic anatomy, a problem arises due to the format of the files being incompatible with current laboratory methodology. Extracting the data from the incompatible files and converting the file to a format that is usable is necessary in order to use the ultrasound files as desired. Figure 1. The progression of the process of creating a computational model from real, imaged anatomy. Image A shows segmented structures in a single ultrasound frame. Image B shows that each frame in the ultrasound sequence is segmented. Image C displays the computational model derived from the images in A and B.