Learning-based techniques have recently been shown to be effective for denoising Monte Carlo rendering methods. However, there remains a quality gap to state-of-the-art handcrafted denoisers. In this paper, we propose...Learning-based techniques have recently been shown to be effective for denoising Monte Carlo rendering methods. However, there remains a quality gap to state-of-the-art handcrafted denoisers. In this paper, we propose a deep residual learning based method that outperforms both state-of-the-art handcrafted denoisers and learning-based denoisers.Unlike the indirect nature of existing learning-based methods(which e.g., estimate the parameters and kernel weights of an explicit feature based filter), we directly map the noisy input pixels to the smoothed output. Using this direct mapping formulation, we demonstrate that even a simple-and-standard ResNet and three common auxiliary features(depth, normal,and albedo) are sufficient to achieve high-quality denoising. This minimal requirement on auxiliary data simplifies both training and integration of our method into most production rendering pipelines. We have evaluated our method on unseen images created by a different renderer. Consistently superior quality denoising is obtained in all cases.展开更多
Flattening filter-free(FFF) beams generated by medical linear particle accelerators(linacs) have recently been used in radiotherapy clinical practice.FFF beams have fundamental physical parameter differences with resp...Flattening filter-free(FFF) beams generated by medical linear particle accelerators(linacs) have recently been used in radiotherapy clinical practice.FFF beams have fundamental physical parameter differences with respect to standard flattening filter(FF) beams,such that the generally used dosimetric parameters and definitions are not always viable.This study investigates dosimetric parameters for use in the quality assurance of FFF beams generated by medical linacs in radiotherapy.The main characteristics of the photon beams are analyzed using specific data generated by a Varian True Beam linac having both FFF and FF beams of 6 and 10 MV(megavolt) energy,respectively.Definitions for dose profile parameters are suggested,starting from the renormalization of the FFF with respect to the corresponding FF beam.From this point,the flatness concept is translated into one of "un-flatness",and other definitions are proposed,maintaining a strict parallelism between FFF and FF parameter concepts.The quality controls used in establishing a quality assurance program when introducing FFF beams into the clinical environment are given,maintaining similarity to those used for standard FF beams,and recommendations for the introduction of FFF beams into clinical radiotherapy application for breast cancer patients are provided as an example for comparison between FFF and FF for dose distribution and coverage for a target volume.Although there are many advantages of using a FFF beam,especially for advanced radiotherapy techniques,there are a few limitations(e.g.,using a relatively higher energy photon beam for stereotactic radiotherapy(SRT),limited speed of current multileaf collimators(MLCs),and off-axis distance-dependent modulation in intensitymodulated radiation therapy(IMRT)) as well as challenges(e.g.,criteria for beam quality evaluation and penumbra,establishment of dosimetry methods,and consequences of photon target burn-up) that need to be addressed for establishing the FFF beam as a viable alternative to the FF beam.展开更多
<strong>Background:</strong><span><span><span style="font-family:""><span style="font-family:Verdana;"> Postoperative irradiation for brain tumor in pregnan...<strong>Background:</strong><span><span><span style="font-family:""><span style="font-family:Verdana;"> Postoperative irradiation for brain tumor in pregnant women is a matter of concern. </span><b><span style="font-family:Verdana;">Aim:</span></b><span style="font-family:Verdana;"> We aimed to assess the safety of radiotherapy for brain tumors in pregnancy. We here report a successful treatment for anaplastic astrocytoma during pregnancy: surgery + postoperative irradiation. We wish to emphasize how we devised irradiation procedure to achieve both therapeutic effectiveness and safety to the fetus/infant. </span><b><span style="font-family:Verdana;">Case Presentation: </span></b><span style="font-family:Verdana;">A 34-year-old pregnant woman suffered with brain anaplastic astrocytoma. Tumor resection under craniotomy was performed with success. We decided to conduct postoperative radiotherapy at 25 weeks of gestation to reduce the risk of recurrence. We used a flattening filter-free volumetric arc therapy (FFF-VMAT) technique, which can achieve lower out-of-field dose than VMAT with a flattening filter or helical tomotherapy. We prescribed 60 Gy over 30 fractions. During actual beam delivery, surface and rectal dose to the patient (mother) were measured. The total fetal dose was estimated at 0.006 - 0.018 Gy, which is under the threshold set by the ICRP. A male healthy infant was born vaginally at the 37th week of pregnancy. The patient (mother) and the infant are healthy at the time of writing.</span><b><span style="font-family:Verdana;"> Conclusion: </span></b><span style="font-family:Verdana;">FFF-VMAT is a good choice for brain tumors during pregnancy</span></span></span></span><span style="font-family:Verdana;">.</span>展开更多
Objective This study aimed to analyze the dosimetric parameters of volumetric modulated arc therapy(VMAT) and dynamic multileaf collimator(DMLC) plans with lumbosacral spine(LS) and pelvic bone marrow(PBM) sparing in ...Objective This study aimed to analyze the dosimetric parameters of volumetric modulated arc therapy(VMAT) and dynamic multileaf collimator(DMLC) plans with lumbosacral spine(LS) and pelvic bone marrow(PBM) sparing in the flattening filter-free(FFF) model.Methods Thirty patients with cervical cancer were selected for analysis.For each patient,four plans with different strategies were generated:VMAT_FF,VMAT_FFF,DMLC_FF,and DMLC_FFF.Dose volumes for organs at risk,conformity index,heterogeneity index(HI),and total number of machine unit(MU) were compared by paired t-test.Result Compared with plans in the FF model,the irradiated dose to LS significantly decreased both VMAT_FFF and DMLC_FFF plans in the FFF model while the MU increased.The mean dose of LS decreased by 107.2 cGy(P<0.05) in the VMAT_FFF plan.In comparison with the VMAT plans in both models of FF and FFF,the DMLC plans reduced the dose volume of 10 Gy by 5.9%(P<0.05),and the mean dose of LS was 189.6-293.1 cGy.The PBM volume receiving 40 Gy showed a decrease of 0.5%-1.1%(P<0.05) as well.The HI increased in the VMAT plans by approximately 0.053 and 0.039.Conclusion The DMLC plans exhibited the best sparing of the PBM and LS in both FF and FFF models and increased the HI.The plans in the FFF model could limit the volume dose of LS but increase the MU.展开更多
基金supported by the Research Grants Council of the Hong Kong Special Administrative Region, under RGC General Research Fund (Project No. CUHK14217516)
文摘Learning-based techniques have recently been shown to be effective for denoising Monte Carlo rendering methods. However, there remains a quality gap to state-of-the-art handcrafted denoisers. In this paper, we propose a deep residual learning based method that outperforms both state-of-the-art handcrafted denoisers and learning-based denoisers.Unlike the indirect nature of existing learning-based methods(which e.g., estimate the parameters and kernel weights of an explicit feature based filter), we directly map the noisy input pixels to the smoothed output. Using this direct mapping formulation, we demonstrate that even a simple-and-standard ResNet and three common auxiliary features(depth, normal,and albedo) are sufficient to achieve high-quality denoising. This minimal requirement on auxiliary data simplifies both training and integration of our method into most production rendering pipelines. We have evaluated our method on unseen images created by a different renderer. Consistently superior quality denoising is obtained in all cases.
文摘Flattening filter-free(FFF) beams generated by medical linear particle accelerators(linacs) have recently been used in radiotherapy clinical practice.FFF beams have fundamental physical parameter differences with respect to standard flattening filter(FF) beams,such that the generally used dosimetric parameters and definitions are not always viable.This study investigates dosimetric parameters for use in the quality assurance of FFF beams generated by medical linacs in radiotherapy.The main characteristics of the photon beams are analyzed using specific data generated by a Varian True Beam linac having both FFF and FF beams of 6 and 10 MV(megavolt) energy,respectively.Definitions for dose profile parameters are suggested,starting from the renormalization of the FFF with respect to the corresponding FF beam.From this point,the flatness concept is translated into one of "un-flatness",and other definitions are proposed,maintaining a strict parallelism between FFF and FF parameter concepts.The quality controls used in establishing a quality assurance program when introducing FFF beams into the clinical environment are given,maintaining similarity to those used for standard FF beams,and recommendations for the introduction of FFF beams into clinical radiotherapy application for breast cancer patients are provided as an example for comparison between FFF and FF for dose distribution and coverage for a target volume.Although there are many advantages of using a FFF beam,especially for advanced radiotherapy techniques,there are a few limitations(e.g.,using a relatively higher energy photon beam for stereotactic radiotherapy(SRT),limited speed of current multileaf collimators(MLCs),and off-axis distance-dependent modulation in intensitymodulated radiation therapy(IMRT)) as well as challenges(e.g.,criteria for beam quality evaluation and penumbra,establishment of dosimetry methods,and consequences of photon target burn-up) that need to be addressed for establishing the FFF beam as a viable alternative to the FF beam.
文摘<strong>Background:</strong><span><span><span style="font-family:""><span style="font-family:Verdana;"> Postoperative irradiation for brain tumor in pregnant women is a matter of concern. </span><b><span style="font-family:Verdana;">Aim:</span></b><span style="font-family:Verdana;"> We aimed to assess the safety of radiotherapy for brain tumors in pregnancy. We here report a successful treatment for anaplastic astrocytoma during pregnancy: surgery + postoperative irradiation. We wish to emphasize how we devised irradiation procedure to achieve both therapeutic effectiveness and safety to the fetus/infant. </span><b><span style="font-family:Verdana;">Case Presentation: </span></b><span style="font-family:Verdana;">A 34-year-old pregnant woman suffered with brain anaplastic astrocytoma. Tumor resection under craniotomy was performed with success. We decided to conduct postoperative radiotherapy at 25 weeks of gestation to reduce the risk of recurrence. We used a flattening filter-free volumetric arc therapy (FFF-VMAT) technique, which can achieve lower out-of-field dose than VMAT with a flattening filter or helical tomotherapy. We prescribed 60 Gy over 30 fractions. During actual beam delivery, surface and rectal dose to the patient (mother) were measured. The total fetal dose was estimated at 0.006 - 0.018 Gy, which is under the threshold set by the ICRP. A male healthy infant was born vaginally at the 37th week of pregnancy. The patient (mother) and the infant are healthy at the time of writing.</span><b><span style="font-family:Verdana;"> Conclusion: </span></b><span style="font-family:Verdana;">FFF-VMAT is a good choice for brain tumors during pregnancy</span></span></span></span><span style="font-family:Verdana;">.</span>
基金General program of National Natural Science Foundation of China,31371425microRNA regulates the mechanism of mitochondrial fusion protein OPA1 involved in apoptosis induced by chemotherapy drugs Special fund of NSFC director,31240025regulation of mitochondrial injury on apoptosis induced by sorafenib
文摘Objective This study aimed to analyze the dosimetric parameters of volumetric modulated arc therapy(VMAT) and dynamic multileaf collimator(DMLC) plans with lumbosacral spine(LS) and pelvic bone marrow(PBM) sparing in the flattening filter-free(FFF) model.Methods Thirty patients with cervical cancer were selected for analysis.For each patient,four plans with different strategies were generated:VMAT_FF,VMAT_FFF,DMLC_FF,and DMLC_FFF.Dose volumes for organs at risk,conformity index,heterogeneity index(HI),and total number of machine unit(MU) were compared by paired t-test.Result Compared with plans in the FF model,the irradiated dose to LS significantly decreased both VMAT_FFF and DMLC_FFF plans in the FFF model while the MU increased.The mean dose of LS decreased by 107.2 cGy(P<0.05) in the VMAT_FFF plan.In comparison with the VMAT plans in both models of FF and FFF,the DMLC plans reduced the dose volume of 10 Gy by 5.9%(P<0.05),and the mean dose of LS was 189.6-293.1 cGy.The PBM volume receiving 40 Gy showed a decrease of 0.5%-1.1%(P<0.05) as well.The HI increased in the VMAT plans by approximately 0.053 and 0.039.Conclusion The DMLC plans exhibited the best sparing of the PBM and LS in both FF and FFF models and increased the HI.The plans in the FFF model could limit the volume dose of LS but increase the MU.
