TY  - JOUR
AB  - Background and purpose: Conventionally, the contours annotated during magnetic resonance-guided radiation therapy (MRgRT) planning are manually corrected during the RT fractions, which is a time-consuming task. Deep learning-based segmentation can be helpful, but the available patient-specific approaches require training at least one model per patient, which is computationally expensive. In this work, we introduced a novel framework that integrates fraction MR volumes and planning segmentation maps to generate robust fraction MR segmentations without the need for patient-specific retraining. Materials and methods: The dataset included 69 patients (222 fraction MRs in total) treated with MRgRT for abdominal cancers with a 0.35 T MR-Linac, and annotations for eight clinically relevant abdominal structures (aorta, bowel, duodenum, left kidney, right kidney, liver, spinal canal and stomach). In the framework, we implemented two alternative models capable of generating patient-specific segmentations using the planning segmentation as prior information. The first one is a 3D UNet with dual-channel input (i.e. fraction MR and planning segmentation map) and the second one is a modified 3D UNet with double encoder for the same two inputs. Results: On average, the two models with prior anatomical information outperformed the conventional population-based 3D UNet with an increase in Dice similarity coefficient >4%. In particular, the dual-channel input 3D UNet outperformed the one with double encoder, especially when the alignment between the two input channels is satisfactory. Conclusion: The proposed workflow was able to generate accurate patient-specific segmentations while avoiding training one model per patient and allowing for a seamless integration into clinical practice.
AU  - De Benetti, F.*
AU  - Delopoulos, N.*
AU  - Belka, C.*
AU  - Corradini, S.*
AU  - Navab, N.*
AU  - Wendler, T.*
AU  - Albarqouni, S.
AU  - Landry, G.*
AU  - Kurz, C.*
C1  - 74691
C2  - 57561
TI  - Enhancing patient-specific deep learning based segmentation for abdominal magnetic resonance imaging-guided radiation therapy: A framework conditioned on prior segmentation.
JO  - Phys. Imag. Radiat. Oncology
VL  - 34
PY  - 2025
SN  - 2405-6316
ER  - 

TY  - JOUR
AB  - BACKGROUND AND PURPOSE: Magnetic resonance imaging-guided radiotherapy (MRgRT) facilitates high accuracy, small margins treatments at the cost of time-consuming and labor-intensive manual delineation of organs-at-risk (OARs). Auto-segmentation models show promise in streamlining this workflow. This study investigates the clinical applicability of a set of thoracic OAR segmentation models for baseline treatment planning in lung tumor patients. We investigate the use of the models for treatment at a 0.35 T MR-linac, assess their potential to reduce physician workload in terms of time savings and quantify the extent of required manual corrections, providing insights into the value of their integration into clinical practice. MATERIALS AND METHODS: Deep-learning based auto-segmentation models for 9 thoracic OARs were integrated into the MRgRT workflow. Two groups of 11 lung cancer cases each were prospectively considered. For Group 1 auto-segmentation contours were corrected by physicians, for Group 2 manual contouring according to standard clinical workflows was performed. Contouring times were recorded for both. Time savings between the groups as well as correlations of the extent of corrections to correction times for Group 1 patients were analyzed. RESULTS: The model performed consistently well across all Group 1 cases. Median contouring times were reduced for six out of nine OARs leading to a reduction of 50.3 % or 12.6 min in median total contouring time. CONCLUSION: Feasibility of auto-segmentation for baseline treatment planning at the 0.35 T MR-linac was shown with significant time savings demonstrated. Time saving potential could not be estimated from model geometric performance metrics.
AU  - Delopoulos, N.*
AU  - Marschner, S.*
AU  - Lombardo, E.*
AU  - Ribeiro, M.F.*
AU  - Rogowski, P.*
AU  - Losert, C.*
AU  - Winderl, T.*
AU  - Albarqouni, S.
AU  - Belka, C.*
AU  - Corradini, S.*
AU  - Kurz, C.*
AU  - Landry, G.*
C1  - 75441
C2  - 58323
CY  - Radarweg 29, 1043 Nx Amsterdam, Netherlands
TI  - Implementation and clinical evaluation of an in-house thoracic auto-segmentation model for 0.35 T magnetic resonance imaging guided radiotherapy.
JO  - Phys. Imag. Radiat. Oncology
VL  - 35
PB  - Elsevier
PY  - 2025
SN  - 2405-6316
ER  - 

TY  - JOUR
AB  - Background and Purpose: Radiotherapy of thoracic tumours can lead to side effects in the lung, which may benefit from early diagnosis. We investigated the potential of X-ray dark-field computed tomography by a proof-of-principle murine study in a clinically relevant radiotherapeutic setting aiming at the detection of radiation-induced lung damage. Material and Methods: Six mice were irradiated with 20 Gy to the entire right lung. Together with five unirradiated control mice, they were imaged using computed tomography with absorption and dark-field contrast before and 16 weeks post irradiation. Mean pixel values for the right and left lung were calculated for both contrasts, and the right-to-left-ratio R of these means was compared. Radiologists also assessed the tomograms acquired 16 weeks post irradiation. Sensitivity, specificity, inter- and intra-reader accuracy were evaluated. Results: In absorption contrast the group-average of R showed no increase in the control group and increased by 7% (p = 0.005) in the irradiated group. In dark-field contrast, it increased by 2% in the control group and by 14% (p = 0.005) in the irradiated group. Specificity was 100% for both contrasts but sensitivity was almost four times higher using dark-field tomography. Two cases were missed by absorption tomography but were detected by dark-field tomography. Conclusions: The applicability of X-ray dark-field computed tomography for the detection of radiation-induced lung damage was demonstrated in a pre-clinical mouse model. The presented results illustrate the differences between dark-field and absorption contrast and show that dark-field tomography could be advantageous in future clinical settings.
AU  - Burkhardt, R.
AU  - Gora, T.*
AU  - Fingerle, A.A.*
AU  - Sauter, A.P.*
AU  - Meurer, F.*
AU  - Gassert, F.T.*
AU  - Dobiasch, S.
AU  - Schilling, D.
AU  - Feuchtinger, A.
AU  - Walch, A.K.
AU  - Multhoff, G.
AU  - Herzen, J.*
AU  - Noel, P.B.*
AU  - Rummeny, E.J.*
AU  - Combs, S.E.
AU  - Schmid, T.E.
AU  - Pfeiffer, F.*
AU  - Wilkens, J.J.*
C1  - 63301
C2  - 51465
SP  - 11-16
TI  - In-vivo X-ray dark-field computed tomography for the detection of radiation-induced lung damage in mice.
JO  - Phys. Imag. Radiat. Oncology
VL  - 20
PY  - 2021
SN  - 2405-6316
ER  - 

TY  - JOUR
AB  - Background and purpose: Proton therapy may be promising for treating non-small-cell lung cancer due to lower doses to the lung and heart, as compared to photon therapy. A reported challenge is degradation, i.e., a smoothing of the depth-dose distribution due to heterogeneous lung tissue. For pencil beams, this causes a distal falloff widening and a peak-to-plateau ratio decrease, not considered in clinical treatment planning systems. Materials and methods: We present a degradation model implemented into an analytical dose calculation, fully integrated into a treatment planning workflow. Degradation effects were investigated on target dose, distal dose falloffs, and mean lung dose for ten patient cases with varying anatomical characteristics. Results: For patients with pronounced range straggling (in our study large tumors, or lesions close to the mediastinum), degradation effects were restricted to a maximum decrease in target coverage (D95 of the planning target volume) of 1.4%. The median broadening of the distal 80–20% dose falloffs was 0.5 mm at the maximum. For small target volumes deep inside lung tissue, however, the target underdose increased considerably by up to 26%. The mean lung dose was not negatively affected by degradation in any of the investigated cases. Conclusion: For most cases, dose degradation due to heterogeneous lung tissue did not yield critical organ at risk overdosing or overall target underdosing. However, for small and deep-seated tumors which can only be reached by penetrating lung tissue, we have seen substantial local underdose, which deserves further investigation, also considering other prevalent sources of uncertainty.
AU  - Winter, J.
AU  - Ellerbrock, M.*
AU  - Jäkel, O.*
AU  - Greilich, S.*
AU  - Bangert, M.*
C1  - 59248
C2  - 48687
SP  - 32-38
TI  - Analytical modeling of depth-dose degradation in heterogeneous lung tissue for intensity-modulated proton therapy planning.
JO  - Phys. Imag. Radiat. Oncology
VL  - 14
PY  - 2020
SN  - 2405-6316
ER  - 

TY  - JOUR
AB  - Background and purpose: Microbeam radiotherapy (MRT) is a preclinical concept in radiation oncology with arrays of alternating micrometer-wide high-dose peaks and low-dose valleys. Experiments demonstrated a superior normal tissue sparing at similar tumor control rates with MRT compared to conventional radiotherapy. Possible clinical applications are currently limited to large third-generation synchrotrons. Here, we investigated the line-focus X-ray tube as an alternative microbeam source. Materials and methods: We developed a concept for a high-voltage supply and an electron source. In Monte Carlo simulations, we assessed the influence of X-ray spectrum, focal spot size, electron incidence angle, and photon emission angle on the microbeam dose distribution. We further assessed the dose distribution of microbeam arc therapy and suggested to interpret this complex dose distribution by equivalent uniform dose. Results: An adapted modular multi-level converter can supply high-voltage powers in the megawatt range for a few seconds. The electron source with a thermionic cathode and a quadrupole can generate an eccentric, high-power electron beam of several 100 keV energy. Highest dose rates and peak-to-valley dose ratios (PVDRs) were achieved for an electron beam impinging perpendicular onto the target surface and a focal spot smaller than the microbeam cross-section. The line-focus X-ray tube simulations demonstrated PVDRs above 20. Conclusion: The line-focus X-ray tube is a suitable compact source for clinical MRT. We demonstrated its technical feasibility based on state-of-the-art high-voltage and electron-beam technology. Microbeam arc therapy is an effective concept to increase the target-to-entrance dose ratio of orthovoltage microbeams.
AU  - Winter, J.
AU  - Galek, M.*
AU  - Matejcek, C.*
AU  - Wilkens, J.J.*
AU  - Aulenbacher, K.*
AU  - Combs, S.E.
AU  - Bartzsch, S.
C1  - 59363
C2  - 48793
SP  - 74-81
TI  - Clinical microbeam radiation therapy with a compact source: Specifications of the line-focus X-ray tube.
JO  - Phys. Imag. Radiat. Oncology
VL  - 14
PY  - 2020
SN  - 2405-6316
ER  -