Unmet Need
Computed tomography (CT) is one of the primary imaging modalities used to image focal liver lesions and assess the efficacy of oncological therapy. Typically, patients are scanned using iodine contrast-enhanced CT, where the change in lesion size over time is an indicator of therapeutic efficacy. However, exclusively relying on lesion size is insufficient for many anti-angiogenic therapies, as such therapies typically result in rapid changes in tumor vasculature within days while size reduction occurs later, in months. Furthermore, current imaging technologies are not sensitive to iodine, so high concentrations are needed, which can be dangerous especially in patients with compromised kidney functions in the event of repeated administration. Alternatively, dual-energy computed tomography (DECT) is a technique that uses two separate x-ray photon energy spectra in order to allow the reconstruction of different images based on varying attenuation properties at different energies. Specifically, it can visualize iodine at lower concentrations, and may provide a more sensitive method of measuring tumor shape, size, and vasculature. Although it potentially addresses the limitations of CT imaging, DECT suffers from bias and large variability in iodine quantitation due to varying patient habitus, imaging protocol, and scanner specifications, in which the effects are especially severe for pediatric and obese patients. Thus, there is a need for a method of decreasing the bias in iodine quantitation in order to use dual-energy systems for oncological imaging.
Technology Overview
Johns Hopkins researchers have developed a method of controlling and minimizing the sources of bias in DECT. First, they have established an end-to-end, patient-specific model for designing dual-energy systems and protocols that are robust against bias by developing a theoretical framework that relates various system parameters to iodine quantitation accuracy and variability. Then, they combined it with an artificial intelligence approach to remove residual bias, by identifying optimal dual-energy acquisition protocols and developing a decomposition algorithm using artificial intelligence-based image reconstruction. Not only does this technology have the potential to optimize the imaging pipeline for integration into oncological imaging, but it also enables the introduction of an iodine biomarker for therapeutic assessment of focal lesions, significantly improving upon the traditional size-dependent measure of treatment efficacy.
Stage of Development
Proof of Concept.
Patent
N/A
Publication
