My team at the University of Texas MD Anderson Cancer Center has developed a multimedia structured reporting system, called ViSion, for diagnostic radiology and other image-based medical disciplines. Attached is a recent article that gives an overview of the sytsem. ViSion operates by allowing a radiologist to capture key images of diagnostic findings (e.g., tumors and metastases), recording a radiologist's voice description of each finding, extracting metadata to tag the images with anatomical location and radiological observation/diagnosis, and assembling a multimedia structured report. Furthermore, ViSion provides a means to link image findings from serial examinations in order to automate disease assessment with graphical disease timelines. ViSion is analogous to being the Facebook of medicine.
Although ViSion's primary purpose is to generate a new style of patient-centric electronic medical record, but one of its added benefits is the ability to perform data mining to match potential subjects with clinical trials. A patient can initiate a search for clinical trials, and ViSion will match that subject's image findings with the inclusion and exclusion criteria from clinical trials that are registered in the system. A patient and/or his/her physician can initiate contact with a prinicipal investigator (PI) for consideration of enrollment into a particular trial. If the PI accepts the patient, then the PI may be granted electronic access to the patient's ViSion data for monitoring during the clinical trial process. ViSion can aggregate data from multiple patients to automate calculation of RECIST and other response criteria, as well as calcuate progression free survival (PFS) and waterfall plots.
We welcome any feedback and invite interest from potential beta sites for evaluation of the ViSion system.