Robust contour propagation using deep learning and image registration for online adaptive proton therapy of prostate cancer.

PURPOSE: To develop and validate a robust and accurate registration pipeline for automatic contour propagation for online adaptive Intensity-Modulated Proton Therapy (IMPT) of prostate cancer using elastix software and deep learning. METHODS: A three-dimensional (3D) Convolutional Neural Network was trained for automatic bladder segmentation of the computed tomography (CT) scans. The automatic bladder segmentation alongside the computed tomography (CT) scan is jointly optimized to add explicit knowledge about the underlying anatomy to the registration algorithm. We included three datasets from different institutes and CT manufacturers. The first was used for training and testing the ConvNet, where the second and the third were used for evaluation of the proposed pipeline. The system performance was quantified geometrically using the dice similarity coefficient (DSC), the mean surface distance (MSD), and the 95% Hausdorff distance (HD). The propagated contours were validated clinically through generating the associated IMPT plans and compare it with the IMPT plans based on the manual delineations. Propagated contours were considered clinically acceptable if their treatment plans met the dosimetric coverage constraints on the manual contours. RESULTS: The bladder segmentation network achieved a DSC of 88% and 82% on the test datasets. The proposed registration pipeline achieved a MSD of 1.29 ± 0.39, 1.48 ± 1.16, and 1.49 ± 0.44 mm for the prostate, seminal vesicles, and lymph nodes, respectively, on the second dataset and a MSD of 2.31 ± 1.92 and 1.76 ± 1.39 mm for the prostate and seminal vesicles on the third dataset. The automatically propagated contours met the dose coverage constraints in 86%, 91%, and 99% of the cases for the prostate, seminal vesicles, and lymph nodes, respectively. A Conservative Success Rate (CSR) of 80% was obtained, compared to 65% when only using intensity-based registration. CONCLUSION: The proposed registration pipeline obtained highly promising results for generating treatment plans adapted to the daily anatomy. With 80% of the automatically generated treatment plans directly usable without manual correction, a substantial improvement in system robustness was reached compared to a previous approach. The proposed method therefore facilitates more precise proton therapy of prostate cancer, potentially leading to fewer treatment-related adverse side effects.

Stereotactic body radiation therapy for liver tumors: impact of daily setup corrections and day-to-day anatomic variations on dose in target and organs at risk.

PURPOSE: To assess day-to-day differences between planned and delivered target volume (TV) and organ-at-risk (OAR) dose distributions in liver stereotactic body radiation therapy (SBRT), and to investigate the dosimetric impact of setup corrections. METHODS AND MATERIALS: For 14 patients previously treated with SBRT, the planning CT scan and three treatment scans (one for each fraction) were included in this study. For each treatment scan, two dose distributions were calculated: one using the planned setup for the body frame (no correction), and one using the clinically applied (corrected) setup derived from measured tumor displacements. Per scan, the two dose distributions were mutually compared, and the clinically delivered distribution was compared with planning. Doses were recalculated in equivalent 2-Gy fraction doses. Statistical analysis was performed with the linear mixed model. RESULTS: With setup corrections, the mean loss in TV coverage relative to planning was 1.7%, compared with 6.8% without corrections. For calculated equivalent uniform doses, these figures were 2.3% and 15.5%, respectively. As for the TV, mean deviations of delivered OAR doses from planning were small (between -0.4 and +0.3 Gy), but the spread was much larger for the OARs. In contrast to the TV, the mean impact of setup corrections on realized OAR doses was close to zero, with large positive and negative exceptions. CONCLUSIONS: Daily correction of the treatment setup is required to obtain adequate TV coverage. Because of day-to-day patient anatomy changes, large deviations in OAR doses from planning did occur. On average, setup corrections had no impact on these doses. Development of new procedures for image guidance and adaptive protocols is warranted.

Deep-ultraviolet quantum interference metrology with ultrashort laser pulses.

Precision spectroscopy at ultraviolet and shorter wavelengths has been hindered by the poor access of narrow-band lasers to that spectral region. We demonstrate high-accuracy quantum interference metrology on atomic transitions with the use of an amplified train of phase-controlled pulses from a femtosecond frequency comb laser. The peak power of these pulses allows for efficient harmonic upconversion, paving the way for extension of frequency comb metrology in atoms and ions to the extreme ultraviolet and soft x-ray spectral regions. A proof-of-principle experiment was performed on a deep-ultraviolet (2 x 212.55 nanometers) two-photon transition in krypton; relative to measurement with single nanosecond laser pulses, the accuracy of the absolute transition frequency and isotope shifts was improved by more than an order of magnitude.