Search

Variations in patient positioning can profoundly influence treatment outcomes, so accurate positioning is crucial for precise radiotherapy dose delivery. This study introduces a novel application of Cherenkov imaging and hypothesizes that it allows for the precise quantification of local-regional tissue deformations to improve the accuracy of patient positioning in breast cancer radiotherapy. For the first time, local-regional deformation is quantified based on Cherenkov imaged vasculature, opening a new avenue to improve the precision of patient positioning. Optical attenuation by blood provides sufficient contrast between vasculature and surrounding tissues. After segmenting blood vessels within Cherenkov images as patient-specific biological fiducial markers, a rigid and non-rigid combined registration was deployed to quantify both inter- and intra-fraction positioning accuracy. The submillimeter accuracy was validated by imaging an anthropomorphic chest phantom with similar human vasculature during a simulated breast radiotherapy treatment, whereas the inter- and intra-fraction variations were simulated by couch shifts and respiratory motion. For 10 patients, 2D maps of local-regional tissue deformations based on non-rigid registration followed by a global shift based on a rigid registration has been quantified in the treatment region for the first time. The accuracy was validated to be within 0.83 ± 0.49 mm for the simulated inter- and intra-fraction variations. A paired t-test revealed no significant difference between the simulated and quantified variations. (P value = 0.2883 > 0.05). A retrospective Cherenkov imaging dataset including 10 breast cancer patients was analyzed for patient positioning variations within their treatment course, revealing an inter-fraction setup uncertainty of 3.71 ± 2.36 mm. Quantitative 2D deformation maps per fraction indicated local-regional deformation in addition to conventional global shifts. Much fewer deformations of 0.0487 ± 0.0385 mm quantified by the non-rigid registration performed after the rigid registration was observed compared to the rigid shift of 3.66 ± 2.35 mm in a paired t-test (P value < 0.0001), which indicates rigid registration captured the majority of global variations, with non-rigid registration addressing the residual local deformations. This study reports the first directly observed local-regional deformation and a method to precisely quantify the global and local variations in patient positioning based on rigid and non-rigid registrations using Cherenkov imaged vasculature. This novel approach demonstrates the feasibility of providing real-time quantitative imaging guidance to inform inter- and intra-fraction positioning, enhancing the precision of breast cancer radiotherapy. [ABSTRACT FROM AUTHOR]

Cherenkov imaging utilizes light emitted during radiation therapy, allowing for visualization of radiation treatments on patients. Cherenkov emissions indicate the treated region and therefore can be used for positional verification; however, there is no method to confirm that the Cherenkov image matches the prediction from the treatment planning system (TPS). In this study, the potential of incorporating a predicted surface dose overlay (PSDO) with Cherenkov image review is examined. PSDOs were generated using the TPS RTPlan and RTDose files for each beam and CT scan. A surface rendering of the patient was created using a non-zero Hounsfield unit threshold. At each point on the surface rendering, the planned dose at the surface is generated by sampling normal to and at a 5mm depth (where most of Cherenkov light is generated) into the dose volume. The PSDO is generated using a 14% isodose of the 5mm planned maximum surface dose and displayed on top of the Cherenkov images. Evaluation of the PSDO was performed on treatments delivered to phantoms at the planned position and after couch shifts to simulate inaccurate patient setups. All patient imaging review was performed on an IRB approved protocol. The patient imaging was scored as congruent if the PSDO shape, size and position visually matched the Cherenkov emissions and non-congruent if not matching. If part of the treatment area was blocked (clothes, sheets or gantry), only the visualized treatment was evaluated. In phantom studies using tangential breast plans, the predicted surface dose overlay was visually congruent at the planned position and non-congruent when the phantom was shifted from the initial position. The predicted plan overlay was less sensitive for treatment plans delivered on the abdomen of the phantom, an area that lacks distinctive anatomical features. In these areas lacking anatomy, accuracy of the technique was restored if biologic fiducials, i.e. blood vessels, were used in conjunction with the PSDO. For patient imaging, 604 treatment fractions for 40 consecutive patients receiving standard of care treatment and pre-treatment imaging were evaluated. Non-congruence of the PSDO with the Cherenkov emission image was detected in 2 treatment courses, totaling 33 of the 604 (5.4%) reviewed treatment fractions. One case was due to anatomy change that occurred during the treatment course. The second case was due to inaccuracy in patient setup and resulted in excess contralateral chest wall dose. This second case was not previously recognized by the treatment team and was not detected by pre-treatment weekly port films or daily SGRT. PSDO incorporated into Cherenkov imaging systems is a useful tool for evaluating accuracy of treatment delivery and has potential for improving treatment quality. Further work is warranted to optimize and determine the added benefit of this technique in large patient studies. [ABSTRACT FROM AUTHOR]

In-vivo beam monitoring will be necessary for safe clinical translation of ultra-high dose rate (UHDR) proton therapy. Although there is no perfect detector for this extreme radiation environment, scintillation imaging dosimetry (SID) has shown promise for accurate dose and dose rate measurements at UHDR timescales. In this study we characterize the noise resolution of an SID system during delivery of modulated UHDR PBS proton, showcasing its utility as an online error-prevention tool. Delivery of complex, 250 MeV UHDR PBS treatment plans with PBS dose rates above 65Gy/s at treatment isocenter were designed with an in-house planning script. A set of pre-defined delivery errors were added to both simple, square-field plans, and complex plans designed for SBRT in the lungs. Spot intensity errors from 1-5% per spot and spot displacements from 1-3 mm were randomly inserted into plans. Surface dose images were collected at 4.5 kHz/0.23 mm resolution during delivery of the planned beams via an intensified CMOS camera facing a block of solid with a top scintillation layer. The average spot-to-spot intensity variations for uniform fields was 2.5+/-0.4% and spot positioning accuracy was 0.49 ± 0.15mm. In uniform fields with planned delivery errors, the SID was able to detect spot MU intensity changes of 5%, measuring 2% differences in dose, which agreed with the planned dose distribution, and spot position changes as low as 1 mm were seen on imaging. For the modulated fields, spot intensity errors as low as 2% were detected, and planned spot position errors of 0.5mm were detected. Of note, imaged dose and dose rate maps showed strong agreement between simulation and imaging, with average gamma passings rate of 98% and 99%, respectively, at 3%/2 mm. In this study, an SID was shown to be able to detect single-spot intensity errors as low as 5%, offering higher sensitivity than gamma analysis. Additionally, the proposed technique allows for monitoring dose-time profiles, potentially alerting the user if an in-delivery error has occurred in real time. [ABSTRACT FROM AUTHOR]

Many modern biomedical technologies rely on detecting in vivo optical signals as a noninvasive means for diagnosis or treatment of disease. The interaction of light in superficial tissue is highly dependent on an individual's skin pigmentation (i.e. epidermal melanin) and must be considered for accurate, quantitative measurements. Cherenkov imaging, an emerging technology that detects light from patients undergoing radiation treatment, could face similar limitations. We present the first reported in vivo Cherenkov imaging study of a diverse patient population towards mitigating the impact of skin pigmentation on quantitative Cherenkov light-based dosimetry. A multi-institutional collaboration was formed to increase the opportunity for imaging a diverse population. Cherenkov imaging was completed with a time-gated, iCMOS camera, and color background images were taken simultaneously with an RGB camera module under standardized lighting. Under an IRB-approved retrospective protocol, skin pigmentation was assessed per patient by calculating the relative luminance (L = 0.2126*R + 0.7152*G + 0.0722*B) of their treated area from the color images. Additionally, 2D dose maps were generated by projecting the exponentially-weighted dose from the surface to 5mm into the body, representative of Cherenkov emission, and used for reference surface dose estimations. To date, N 6MV =23 and N 15MV =20 imaged patients, encompassing a wide variety of skin pigmentations, fit our assessment criteria: right-sided breast radiotherapy patients without reconstructive implants or temporary expanders. The initial generated intensity of Cherenkov light is proportional to the radiation dose delivered, providing a unique opportunity for non-contact, real-time surface dosimetry. However, the detected light emission is highly dependent on patients' skin pigmentation. Our results revealed that, for the same dose, the Cherenkov emission intensity was nearly four times less for dark skin patients compared to their Caucasian counterparts. Additionally, a linear relationship exists between Cherenkov intensity per unit dose and individuals' relative skin luminance, enabling a linear calibration factor based on skin pigmentation. Application of this calibration factor significantly improved Cherenkov-to-dose linearity amongst the cohort, from R2=0.79 to 0.96 for 6MV and R2=0.19 to 0.91 for 15MV (p<0.05), marking the highest reported linearity for in vivo patient data. This study is the first of its kind dedicated to assessing Cherenkov imaging amongst a diverse patient population, and developed a linear calibration technique based upon measuring the skin luminance for each patient. It demonstrates significant mitigation of the effect of skin pigmentation through the addition of color imaging, representing a critical step towards achieving quantitative Cherenkov dosimetry. [ABSTRACT FROM AUTHOR]

FLASH radiotherapy based on ultra-high dose rate (UHDR) is actively being studied by the radiotherapy community. Dedicated UHDR electron devices are currently a mainstay for FLASH studies. This work is to present the dosimetry and treatment planning tools for the UHDR capable compact electron Linac for preclinical research and FLASH-radiotherapy (RT) clinical trials. Film based dosimetry methodology has been established for the commissioning of UHDR electron Linac. Monte Carlo (MC) beam model for treatment planning was configured and validated with a simulation toolkit. The geometry and electron source characteristics, such as energy spectrum and beamline parameters, were tuned to match the central-axis depth dose (PDD) and lateral profiles for the pristine beam measured during machine commissioning. Diode-detector (UHDR capable) based beam monitoring method was implemented. To facility fast and accurate PDD measurements, MC model of the diode detector was configured, and its electron energy response was investigated. Correction method for the diode reading to facilize UHDR PDD measurements was established for routine use during preclinical research. Preclinical research acceptable treatment planning method was established. Good agreement between the MC beam model and commissioning data were demonstrated with maximal discrepancy < 3% for PDDs and profiles. 100% gamma pass rate was achieved for all PDDs and profiles with the criteria of 2mm/3%. With the criteria of 2mm/2%, maximum, minimum and mean gamma pass rates were (100.0%, 93.8%, 98.7%) for PDDs and (100.0%, 96.7%, 99.4%) for profiles, respectively. Both film and diode detector showed good agreements in PDD measurements for UHDR electron beam. A validated MC beam model for the treatment planning of UHDR capable compact Linac is presented for the first time. The beam model presented in this work should facilitate translational and clinical FLASH-RT for trials conducted on the compact UHDR electron platform. [ABSTRACT FROM AUTHOR]

Purpose: This study investigates scintillation dosimetry coupled with Cherenkov imaging for in vivo dose monitoring during whole breast radiation therapy (WBRT). Given recent observations of excess dose to the contralateral breast (CB), in vivo dosimetry (IVD) could help ensure accurate dose delivery and decrease risks of secondary cancer. This work presents a rapid, streamlined alternative to traditional IVD, providing direct visualization of measurement location relative to the treatment field on the patient., Methods and Materials: Ten WBRT patients consented under an institutional review board-approved protocol were monitored with scintillation dosimetry and always-on Cherenkov imaging, on both their treated and CB for 1 to 3 fractions. Scintillator dosimeters, small plastic discs 1 mm thick and 15 mm in diameter, were calibrated against optically stimulated luminescent dosimeters (OSLDs) to generate an integral output-to-dose conversion, where integral output is measured in postprocessing through a custom fitting algorithm. The discs have been extensively characterized in a previous study for various treatment conditions including beam energy and treatment geometry., Results: A total of 44 dosimetry measurements were evaluated, including 22 treated breast and 22 CB measurements. After integral output-to-dose calibration, in vivo scintillator dosimeters exhibited high linearity (R 2 = 0.99) with paired OSLD readings across all patients. The difference between scintillation and OSLD dose measurements averaged 2.8% of the prescribed dose, or an absolute dose difference of approximately 7 cGy., Conclusions: Integration of scintillation dosimetry with Cherenkov imaging offers an accurate, rapid alternative for in vivo dose verification in WBRT, circumventing the limitations of conventional point dosimeters. The additional benefit of visualizing measurement locations relative to the treatment field provides users an enhanced understanding of results and allows for detection of high dose gradients. Future work will explore the applicability of this technique across a broader range of radiation therapy treatments, aiming to streamline IVD practices., (Copyright © 2024 Elsevier Inc. All rights reserved.)

Purpose: To retrospectively validate the dose and dose rates delivered in FAST-01 clinical trial fields via submillimeter spatial and <0.25 ms temporal resolution scintillation imaging., Methods and Materials: An ultrafast intensified CMOS camera (4.5-12 kHz sampling rate) imaged the light response of a scintillator sheet at the treatment isocenter and irradiated by pencil beam scanning proton fields, including FAST-01 clinical fields. Dose and dose rate linearity studies were performed, followed by camera calibration via gafchromic (external beam therapy 3) film. An EDGE diode detector was placed directly under the scintillator at the surface and at a 5 cm depth for comparison with the imaging data and log files. Frame-by-frame analysis of image stacks yielded dose and dose rate maps for each delivery at the surface. Using the percent depth dose curves and 3-dimensional spot profiles, the surface images were projected to a 5 cm depth for comparison with a secondary diode and log file recordings., Results: Camera response was linear with dose (R 2 = 0.9998) and beam current (R 2 = 0.9883) from 2 to 12 Gy and 20 to 210 nA, respectively. Gamma analysis of the cumulative dose maps at 3%/2 mm indicated a mean passing rate of 100% compared with film. Total irradiation time agreed with the log file recordings with an average deviation of 0.20 ± 0.07 ms. At the surface, the average imaged dose rate across the 7 fields was 114 ± 1 Gy/s, agreeing with the diode within 1% ± 1%. The dose at a 5 cm depth from the projected images (mean, 8.4 Gy) agreed with the reported dose (log files) within 0.13 ± 0.03 Gy (2% ± 1%). The average dose rate from projected images at a 5 cm depth was 62 ± 1 Gy/s, which agreed with the reported and diode values within 2% ± 3%., Conclusions: This study provides the first independent validation of dose and dose rate for clinical proton FAST-01 fields at unprecedented spatiotemporal resolution. Owing to the nontrivial dose rate distributions in pencil beam scanning fields, such direct 2-dimensional dose rate mapping will be important in future pretreatment plan quality assurance., (Published by Elsevier Inc.)

Purpose: Large, rapid extracellular oxygen transients (ΔpO 2 ) have been measured in vivo during ultra-high dose rate radiation therapy; however, it has been unclear if they match intracellular oxygen levels. Here, the endogenously produced protoporphyrin IX (PpIX) delayed fluorescence signal was measured as an intracellular in-vivo oxygen sensor to quantify these transients, with direct comparison to extracellular pO 2 . Intracellular ΔpO 2 is closer to the cellular DNA, the site of major radiobiological damage, and therefore should help elucidate radiochemical mechanisms of the FLASH effect and potentially be translated to human tissue measurement., Methods and Materials: PpIX was induced in mouse skin through intraperitoneal injection of 250 mg/kg of aminolevulinic acid. The animals were also administered a 50 µL intradermal injection of 10 µM oxyphor G4 (PdG4) for phosphorescence lifetime pO 2 measurement. Paired oxygen transients were quantified in leg or flank tissues while delivering 10 MeV electrons in 3 µs pulses at 360 Hz for a total dose of 10 to 28 Gy., Results: Transient reductions in pO 2 were quantifiable in both PpIX delayed fluorescence and oxyphor phosphorescence, corresponding to intracellular and extracellular pO 2 values, respectively. Reponses were quantified for 10, 22, and 28 Gy doses, with ΔpO 2 found to be proportional to the dose on average. The ΔpO 2 values were dependent on initial pO 2 in a logistic function. The average and standard deviations in ΔpO 2 per dose were 0.56 ± 0.18 mm Hg/Gy and 0.43 ± 0.06 mm Hg/Gy for PpIX and oxyphor, respectively, for initial pO 2 > 20 mm Hg. Although there was large variability in the individual animal measurements of ΔpO 2 , the average values demonstrated a direct and proportional correlation between intracellular and extracellular pO 2 changes, following a linear 1:1 relationship., Conclusions: A fundamentally new approach to measuring intracellular oxygen depletion in living tissue showed that ΔpO 2 transients seen during ultra-high dose rate radiation therapy matched those quantified using extracellular oxygen measurement. This approach could be translated to humans to quantify intracellular ΔpO 2 . The measurement of these transients could potentially allow the estimation of intracellular reactive oxygen species production., (Copyright © 2024 Elsevier Inc. All rights reserved.)

Purpose: In this study, a C-series linear accelerator was configured to enable rapid and reliable conversion between the production of conventional electron beams and an ultrahigh-dose-rate (UHDR) electron beamline to the treatment room isocenter for FLASH radiation therapy. Efforts to tune the beam resulted in a consistent, stable UHDR beamline., Methods and Materials: The linear accelerator was configured to allow for efficient switching between conventional and modified electron output modes within 2 minutes. Additions to the air system allow for retraction of the x-ray target from the beamline when the 10 MV photon mode is selected. With the carousel set to an empty port, this grants access to the higher current pristine electron beam normally used to produce clinical photon fields. Monitoring signals related to the automatic frequency control system allows for tuning of the waveguide while the machine is in a hold state so a stable beam is produced from the initial pulse. A pulse counting system implemented on an field-programmable gate array-based controller platform controls the delivery to a desired number of pulses. Beam profiles were measured with Gafchromic film. Pulse-by-pulse dosimetry was measured using a custom electrometer designed around the EDGE diode., Results: This method reliably produces a stable UHDR electron beam. Open-field measurements of the 16-cm full-width, half-maximum gaussian beam saw average dose rates of 432 Gy/s at treatment isocenter. Pulse overshoots were limited and ramp up was eliminated. Over the last year, there have been no recorded incidents that resulted in machine downtime due to the UHDR conversions., Conclusions: Stable 10 MeV UHDR beams were generated to produce an average dose rate of 432 Gy/s at the treatment room isocenter. With a reliable pulse-counting beam control system, consistent doses can be delivered for FLASH experiments with the ability to accommodate a wide range of field sizes, source-to-surface distances, and other experimental apparatus that may be relevant for future clinical translation., (Copyright © 2024 Elsevier Inc. All rights reserved.)

Purpose: To present a Monte Carlo (MC) beam model and its implementation in a clinical treatment planning system (TPS, Varian Eclipse) for a modified ultrahigh dose-rate electron FLASH radiation therapy (eFLASH-RT) linear accelerator (LINAC) using clinical accessories and geometry., Methods and Materials: The gantry head without scattering foils or targets, representative of the LINAC modifications, was modeled in the Geant4-based GAMOS MC toolkit. The energy spectrum (σ E ) and beam source emittance cone angle (θ cone ) were varied to match the calculated open-field central-axis percent depth dose (PDD) and lateral profiles with Gafchromic film measurements. The beam model and its Eclipse configuration were validated with measured profiles of the open field and nominal fields for clinical applicators. An MC forward dose calculation was conducted for a mouse whole-brain treatment, and an eFLASH-RT plan was compared with a conventional (Conv-) RT electron plan in Eclipse for a human patient with metastatic renal cell carcinoma., Results: The eFLASH beam model agreed best with measurements at σ E  = 0.5 MeV and θ cone  = 3.9° ± 0.2°. The model and its Eclipse configuration were validated to clinically acceptable accuracy (the absolute average error was within 1.5% for in-water lateral, 3% for in-air lateral, and 2% for PDDs). The forward calculation showed adequate dose delivery to the entire mouse brain while sparing the organ at risk (lung). The human patient case demonstrated the planning capability with routine accessories to achieve an acceptable plan (90% of the tumor volume receiving 95% and 90% of the prescribed dose for eFLASH and Conv-RT, respectively)., Conclusions: To our knowledge, this is the first functional beam model commissioned in a clinical TPS for eFLASH-RT enabling planning and evaluation with minimal deviation from the Conv-RT workflow. It facilitates the clinical translation because eFLASH-RT and Conv-RT plan quality were comparable for a human patient involving complex geometries and tissue heterogeneity. The methods can be expanded to model other eFLASH irradiators with different beam characteristics., (Copyright © 2021 Elsevier Inc. All rights reserved.)

Purpose: Delivery of radiation at ultrahigh dose rates (UHDRs), known as FLASH, has recently been shown to preferentially spare normal tissues from radiation damage compared with tumor tissues. However, the underlying mechanism of this phenomenon remains unknown, with one of the most widely considered hypotheses being that the effect is related to substantial oxygen depletion upon FLASH, thereby altering the radiochemical damage during irradiation, leading to different radiation responses of normal and tumor cells. Testing of this hypothesis would be advanced by direct measurement of tissue oxygen in vivo during and after FLASH irradiation., Methods and Materials: Oxygen measurements were performed in vitro and in vivo using the phosphorescence quenching method and a water-soluble molecular probe Oxyphor 2P. The changes in oxygen per unit dose (G-values) were quantified in response to irradiation by 10 MeV electron beam at either UHDR reaching 300 Gy/s or conventional radiation therapy dose rates of 0.1 Gy/s., Results: In vitro experiments with 5% bovine serum albumin solutions at 23°C resulted in G-values for oxygen consumption of 0.19 to 0.21 mm Hg/Gy (0.34-0.37 μM/Gy) for conventional irradiation and 0.16 to 0.17 mm Hg/Gy (0.28-0.30 μM/Gy) for UHDR irradiation. In vivo, the total decrease in oxygen after a single fraction of 20 Gy FLASH irradiation was 2.3 ± 0.3 mm Hg in normal tissue and 1.0 ± 0.2 mm Hg in tumor tissue (P < .00001), whereas no decrease in oxygen was observed from a single fraction of 20 Gy applied in conventional mode., Conclusions: Our observations suggest that oxygen depletion to radiologically relevant levels of hypoxia is unlikely to occur in bulk tissue under FLASH irradiation. For the same dose, FLASH irradiation induces less oxygen consumption than conventional irradiation in vitro, which may be related to the FLASH sparing effect. However, the difference in oxygen depletion between FLASH and conventional irradiation could not be quantified in vivo because measurements of oxygen depletion under conventional irradiation are hampered by resupply of oxygen from the blood., (Published by Elsevier Inc.)

Purpose: In this study, procedures were developed to achieve efficient reversible conversion of a clinical linear accelerator (LINAC) and deliver ultrahigh-dose-rate (UHDR) electron or conventional beams to the treatment room isocenter for FLASH radiation therapy., Methods and Materials: The LINAC was converted to deliver UHDR beam within 20 minutes by retracting the x-ray target from the beam's path, positioning the carousel on an empty port, and selecting 10 MV photon beam energy in the treatment console. Dose rate surface and depth dose profiles were measured in solid water phantom at different field sizes with Gafchromic film and an optically stimulated luminescent dosimeter (OSLD). A pulse controller counted the pulses via scattered radiation signal and gated the delivery for a preset pulse count. A fast photomultiplier tube-based Cherenkov detector measured the per pulse beam output at a 2-ns sampling rate. After conversion back to clinical mode, conventional beam output, flatness, symmetry, field size, and energy were measured for all clinically commissioned energies., Results: The surface average dose rates at the isocenter for 1-cm diameter and 1.5-in diameter circular fields and for a jaws-wide-open field were 238 ± 5 Gy/s, 262 ± 5 Gy/s, and 290 ± 5 Gy/s, respectively. The radial symmetry of the beams was within 2.4%, 0.5%, and 0.2%, respectively. The doses from simultaneous irradiation of film and OSLD were within 1%. The photomultiplier tube showed the LINAC required ramp up time in the first 4 to 6 pulses before the output stabilized, after which its stability was within 3%., Conclusions: At the isocenter of the treatment room, 10 MeV UHDR beams were achieved. The beam output was reproducible but requires further investigation of the ramp up time, equivalent to ∼1 Gy, requiring dose monitoring. The UHDR beam can irradiate both small and large subjects to investigate potential FLASH radiobiological effects in minimally modified clinical settings, and the dose rate can be further increased by reducing the source-to-surface distance., (Copyright © 2021 Elsevier Inc. All rights reserved.)

Purpose: The value of Cherenkov imaging as an on-patient, real-time, treatment delivery verification system was examined in a 64-patient cohort during routine radiation treatments in a single-center study., Methods and Materials: Cherenkov cameras were mounted in treatment rooms and used to image patients during their standard radiation therapy regimen for various sites, predominantly for whole breast and total skin electron therapy. For most patients, multiple fractions were imaged, with some involving bolus or scintillators on the skin. Measures of repeatability were calculated with a mean distance to conformity (MDC) for breast irradiation images., Results: In breast treatments, Cherenkov images identified fractions when treatment delivery resulted in dose on the contralateral breast, the arm, or the chin and found nonideal bolus positioning. In sarcoma treatments, safe positioning of the contralateral leg was monitored. For all 199 imaged breast treatment fields, the interfraction MDC was within 7 mm compared with the first day of treatment (with only 7.5% of treatments exceeding 3 mm), and all but 1 fell within 7 mm relative to the treatment plan. The value of imaging dose through clear bolus or quantifying surface dose with scintillator dots was examined. Cherenkov imaging also was able to assess field match lines in cerebral-spinal and breast irradiation with nodes. Treatment imaging of other anatomic sites confirmed the value of surface dose imaging more broadly., Conclusions: Daily radiation therapy can be imaged routinely via Cherenkov emissions. Both the real-time images and the posttreatment, cumulative images provide surrogate maps of surface dose delivery that can be used for incident discovery and/or continuous improvement in many delivery techniques. In this initial 64-patient cohort, we discovered 6 minor incidents using Cherenkov imaging; these otherwise would have gone undetected. In addition, imaging provides automated, quantitative metrics useful for determining the quality of radiation therapy delivery., (Copyright © 2020 Elsevier Inc. All rights reserved.)

Purpose: Patients have reported sensations of seeing light flashes during radiation therapy, even with their eyes closed. These observations have been attributed to either direct excitation of retinal pigments or generation of Cherenkov light inside the eye. Both in vivo human and ex vivo animal eye imaging was used to confirm light intensity and spectra to determine its origin and overall observability., Methods and Materials: A time-gated and intensified camera was used to capture light exiting the eye of a patient undergoing stereotactic radiosurgery in real time, thereby verifying the detectability of light through the pupil. These data were compared with follow-up mechanistic imaging of ex vivo animal eyes with thin radiation beams to evaluate emission spectra and signal intensity variation with anatomic depth. Angular dependency of light emission from the eye was also measured., Results: Patient imaging showed that light generation in the eye during radiation therapy can be captured with a signal-to-noise ratio of 68. Irradiation of ex vivo eye samples confirmed that the spectrum matched that of Cherenkov emission and that signal intensity was largely homogeneous throughout the entire eye, from the cornea to the retina, with a slight maximum near 10 mm depth. Observation of the signal external to the eye was possible through the pupil from 0° to 90°, with a detected emission near 2500 photons per millisecond (during peak emission of the ON cycle of the pulsed delivery), which is over 2 orders of magnitude higher than the visible detection threshold., Conclusions: By quantifying the spectra and magnitude of the signal, we now have direct experimental observations that Cherenkov light is generated in the eye during radiation therapy and can contribute to perceived light flashes. Furthermore, this technique can be used to further study and measure phosphenes in the radiation therapy clinic., (Copyright © 2019 The Author(s). Published by Elsevier Inc. All rights reserved.)