Validation Study of a Gastrointestinal Motility Device With a Radiation-Sparing Protocol

Challenge

A developer of an innovative device for evaluating gastrointestinal motility required a validation study comparing the device's output against an imaging reference standard. Conventional transit studies rely on serial radiographs or fluoroscopy, and a schedule dense enough to validate the device could expose healthy volunteers and patients to meaningful cumulative radiation.

The sponsor needed a protocol that produced an adequate reference standard while limiting radiation exposure, and a core lab able to design the imaging schedule, read the studies consistently and process the data on a timeline compatible with a device development program.

Intrinsic Imaging’s Solution

Intrinsic Imaging's Chief Medical Officer co-authored the protocol design with the sponsor, structuring the imaging schedule to capture the transit timepoints needed for validation while limiting the number and dose of radiographic acquisitions. Low-dose techniques and defined collimation were specified in the charter, and sites were qualified and trained on the reduced-dose protocol.

Radiologists read the reference-standard imaging under a charter that defined marker counting, regional transit scoring and the handling of overlapping anatomy, with a second reader on a prespecified sample to establish reproducibility. Device output was compared against the imaging reference at matched timepoints.

Images and device data were processed through the company's cloud-based platform, linking each subject's imaging timepoints with the corresponding device measurements and delivering an analysis-ready dataset.

Intrinsic Imaging’s Results

• Radiation-sparing imaging schedule reduced per-subject radiographic acquisitions by roughly 40% relative to a conventional transit protocol while preserving the reference standard
• 96 subjects imaged at 4 sites with all reference-standard timepoints evaluable
• Reader reproducibility of regional transit scoring established on a 20% double-read sample
• Imaging reference and device measurements delivered as a matched, analysis-ready dataset through the cloud-based processing platform