Rad AI Showcases Simulation Work on Photon-Counting CT Detector Charge Sharing

According to a recent LinkedIn post from Rad AI, the company has conducted a detailed simulation study on charge sharing in photon-counting detectors for computed tomography applications. The post describes the use of CERN’s Allpix-squared framework to model a realistic detector geometry and operating conditions relevant to medical imaging.

The LinkedIn post highlights that charge sharing was observed in 41.1% of simulated photon interactions, with 58.9% isolated single-pixel events and 32.3% adjacent bipixel sharing. The remaining interactions were split among tripixel, diagonal bipixel, and quadpixel events, aligning with expected Gaussian charge cloud diffusion behavior.

The post suggests that charge sharing remains a significant physical limitation for photon-counting CT, as it can degrade energy resolution and introduce spectral distortion, potentially impacting diagnostic image quality. By quantifying these effects, Rad AI appears to be positioning itself as a technically sophisticated player in photon-counting and spectral CT detector research.

For investors, this type of research-focused content may indicate ongoing investment in advanced simulation capabilities and domain expertise rather than an immediate commercial product. However, progress in understanding and mitigating charge sharing could enhance Rad AI’s credibility with OEM partners, research institutions, and healthcare providers in a market increasingly focused on next-generation CT technologies.

If Rad AI can leverage this work into improved detector designs, algorithms, or software for spectral correction, it may open up future revenue opportunities in high-end imaging systems and AI-driven image quality optimization. The engagement invitation at the end of the post, asking others to share their observed charge sharing fractions, also points to an effort to build technical community ties that could support collaborations and longer-term business development.