Tackling the Real Problem: The Dosimetric Unknowns in Patient Setup

Treatment planning is performed on a static CT dataset, capturing a single snapshot in time. The resulting plan is then delivered to the patient over multiple weeks, relying on efforts to recreate that initial anatomy and setup for every fraction. 

Surface Guided Radiation Therapy (SGRT) technologies have become invaluable in helping position patients consistently, using surface metrics to guide alignment. Yet, typical SGRT tolerances — often set generically at 3–5 mm and 3 degrees — are rarely customized to individual patients or specific anatomical sites. These thresholds don't reflect the actual dosimetric consequences of deviations for a given patient. 

Patients with limited mobility introduce additional challenges in reproducibility. Even with careful setup, small positional inconsistencies can result in subtle, unintended dose shifts. In some cases, evidence of excessive or misplaced dose to normal tissues is only discovered after treatment has been delivered, when the opportunity to intervene has passed. 

Patients, after all, are not phantoms, and Patient Specific Quality Assurance (PSQA) activities lose clinical meaning when a surrogate phantom stands in for the patient for independent secondary calculations and IMRT QA process. After all the checks and approvals are done and a patient is prepared for the start of treatment, their anatomy has already changed — and it continues to change throughout the course of treatment. Breast tissue can swell or develop seromas. Abdominal organs fluctuate with varying gas or solid content. The pelvis sees daily shifts with bladder or rectal filling. Even with Image Guided Radiation Therapy (IGRT) to align to internal anatomy and minimize gross setup errors, the insight into how day-to-day anatomical changes impact the actual dose delivered is still lacking. So the question remains:

• How do we adapt and personalize our SGRT and IGRT setup tolerances based on the true dosimetric situation for each patient?
• How does moving to more individualized tolerances help clinicians decide when it’s safe to treat versus when it’s better to adapt or replan? 

The Challenge Is Recognized — But Not Universally Solved 

Unseen dosimetric uncertainty is now acknowledged across the field. Sophisticated platforms like the Elekta Unity MR-linac offer online adaptive radiotherapy workflows such as Adapt to Position (ATP), which aligns the day’s image to the reference via rigid registration and updates the plan accordingly. 

But these solutions require significant infrastructure and cost. Many clinics lack such technology yet still face the same challenges of anatomy-driven dose variability. 

Where RadCalc Is Heading 

At LAP, we’re addressing this gap by building on our core strengths: precise alignment and independent dose verification. LAP lasers and LUNA 3D offer continuous, non-invasive surface tracking from simulation through delivery. Paired with RadCalc’s 3D QA suite, we’re creating an integrated workflow to make dosimetric insight both patient-specific and actionable. 

RadCalc’s EPID dosimetry already allows clinics to assess delivered dose retrospectively. But what’s needed now is proactive insight — before or during treatment, not just after. 

Redefining QA: Integrated SGRT and Dose Evaluation Workflows 

RadCalc’s future centers on combining SGRT and adaptive QA through three key workflows: 

Pre-Treatment Dose Evaluation 
RadCalc uses a synthetic CT built from the planning CT, updated with LUNA 3D surface data and radiographic imaging, to recalculate dose before beam-on. Console-side dose difference metrics and DVHs help therapists make informed decisions quickly, while physicists can review remotely. 

Intra-Treatment QA 
With LUNA 3D’s live tracking and RadCalc’s dose engine, dose can be monitored during delivery. If patient motion exceeds dosimetric thresholds — even within setup tolerances — alerts or beam pauses can prevent deviations. 

Post-Fraction Analysis 
By combining EPID measurements with accumulated dose over fractions, RadCalc builds a complete picture of what was actually delivered. These insights guide offline adaptation and feed BED-based tools for summing across courses or sites. 

A Smarter, Patient-Centric QA Standard 

Together, these tools help clinics move from fixed tolerances to a personalized, dosimetrically grounded approach — no MR-linac required. With RadCalc, QA becomes dynamic, data-driven, and built around the patient, not the plan. 

Let’s connect and explore how this evolution in QA can support your clinic’s vision. 

Note: German Patent Application No. 10 2023 115  102.9, PCT Patent Application No. EP2024/065566, US Patent Application No. 19/144,366

EFOMP Article.