A convolutional neural network approach for IMRT dose distribution prediction in prostate cancer patients

Kajikawa, Tomohiro; Kadoya, Noriyuki; Ito, Kengo; Takayama, Yoshiki; Chiba, Takahito; Tomori, Seiji; Nemoto, Hitoshi; Dobashi, Suguru; Takeda, Ken; Jingu, Keiichi

doi:10.1093/jrr/rrz051

Cited by 46 publications

(59 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The difficult-to-return zone of Tomioka Town was surveyed using a car-borne survey system, Radi-probe (Chiyoda Technology Corp., Tokyo, Japan. The handheld radiation detector model: HDS-101GN, Mirion Technologies, Inc., Japan) 6,39 . The Radi-probe system was installed in a vehicle and the meter's detector was set on the front passenger seat about 1 m above the ground.…”

Section: Survey Of Ambient Rates and Radionuclidesmentioning

confidence: 99%

“…Real-time maps with color-scaled ambient dose rates and gamma-ray energy spectra can be output. The detected energy peaks of radiocesium ( 134 Cs and 137 Cs) registered in the nuclear library (i.e., detected net count values) and their associated confidence intervals were obtained for the region of interest (with levels 1-10 used as reference values) 7,39 .…”

Section: Survey Of Ambient Rates and Radionuclidesmentioning

confidence: 99%

See 1 more Smart Citation

Environmental Remediation of the difficult-to-return zone in Tomioka Town, Fukushima Prefecture

Cui

Taira

Matsuo

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Temporal variations in ambient dose rates in a restricted area designated as "difficult-to-return" for residents of Tomioka Town, Fukushima Prefecture were evaluated in a car-borne survey during 2018-2019. The median dose rates in the "Decontaminated area" in the difficult-to-return zone decreased rapidly from 1.0 μSv/h to 0.32 μSv/h; however, the median dose rates in the "Non-decontaminated area" and "Radioactive waste storage area" fluctuated between 1.1-1.4 μSv/h and 0.46-0.61 μSv/h, respectively. The detected rate of the cesium-137 (137 Cs) (137 Cs-detected points per all measuring points) in the "Decontaminated area" also decreased rapidly from 64% to 6.7%, accompany with decreasing in ambient dose rates. On the other hand, the detection of 137 Cs in the "Radioactive waste storage area" and "Non-decontaminated area" decreased from 53% to 17% and 93% to 88%, respectively. We confirmed that the dose rates in the Decontaminated area dramatically decreased due to decontamination work aiming to help residents return home. Moreover, the estimated external exposure dose of workers during the present survey was 0.66 mSv/y in the Decontaminated area and 0.55 mSv/y in the Radioactive waste storage area, respectively. This case of Tomioka Town within the "difficult-to-return zone" may be the first reconstruction model for evaluating environmental contamination and radiation exposure dose rates due to artificial radionuclides derived from the nuclear disaster. The Great East Japan Earthquake (magnitude 9.0) and subsequent tsunami on March 11, 2011 caused an accident at the Fukushima Daiichi Nuclear Power Station (FDNPS) that resulted in various radionuclides including cesium-134 (134 Cs), cesium-137 (137 Cs) and iodine-131(131 I) being released into the atmosphere and eventually depositing on land and at sea in the surrounding areas 1. The United Nations Scientific Committee on the Effects of Atomic Radiation estimated the total release of 134 Cs (half-life: 2.1 y), 137 Cs (half-life: 30 y) and 131 I (half-life: 8 d) to be 9.0, 8.8 and 120.0 petabecquerels (PBq), respectively 1. The Japanese government, municipalities and private companies have carried out environmental and individual radiation monitoring programs to confirm the radiation levels in the affected areas 2,3. More than 8 years have passed since the accident and it has been confirmed that artificial radionuclides with a relatively long half-life such as 134 Cs and 137 Cs still exist in soil and plant samples collected around the FDNPS 1-3. Environmental monitoring in Fukushima Prefecture have been carried out by many methods (the airborne survey by monitor stations and personnel, vehicle-borne survey, aerial-vehicle survey and radionuclide analysis of the environmental samples such as soils, sediments and foodstuffs) 3-9. These surveys and the collected data are extremely important to precise evaluation of environmental remediation in the affected areas. Following the FDNPS accident, residential areas, farmlands, forests close to residential areas, and roads...

show abstract

Section: Survey Of Ambient Rates and Radionuclidesmentioning

confidence: 99%

Section: Survey Of Ambient Rates and Radionuclidesmentioning

confidence: 99%

Environmental Remediation of the difficult-to-return zone in Tomioka Town, Fukushima Prefecture

Cui

Taira

Matsuo

et al. 2020

Sci Rep

View full text Add to dashboard Cite

show abstract

“…However, as illustrated in Fig. 10, we have shown that Model I can predict within roughly 2% mean dose error of the prescription dose for each structure for any beam number setup, which is competitive to any other deep learning‐based dose prediction method in literature, 47,48,50–52,54,55,73,74 most of which use only one type of beam geometry setup for their study. The benefit of training such a beam‐flexible model is that, during deployment, the treatment planner may now adjust the beam number in real time and possibly find a beam configuration the same or better plan quality as before.…”

Section: Discussionmentioning

confidence: 79%

“…The development of the fully convolutional network (FCN) 46 allowed for pixel‐wise prediction using supervised learning, which opened the door for voxel‐wise dose prediction and generation of DVH curves in treatment planning. Recently, many researchers have developed different deep learning models for predicting clinical dose distributions for IMRT and volumetric‐modulated arc therapy (VMAT) modalities on different treatment sites such as lung, prostate, and head‐and‐neck 47–53 . However, all of these models used static beam orientations for their study, thus limiting their uses in the treatment planning workflow to a subset of common treatment plans based on protocol.…”

Section: Introductionmentioning

confidence: 99%

Using deep learning to predict beam‐tunable Pareto optimal dose distribution for intensity‐modulated radiation therapy

et al. 2020

View full text Add to dashboard Cite

Purpose: Many researchers have developed deep learning models for predicting clinical dose distributions and Pareto optimal dose distributions. Models for predicting Pareto optimal dose distributions have generated optimal plans in real time using anatomical structures and static beam orientations. However, Pareto optimal dose prediction for intensity-modulated radiation therapy (IMRT) prostate planning with variable beam numbers and orientations has not yet been investigated. We propose to develop a deep learning model that can predict Pareto optimal dose distributions by using any given set of beam angles, along with patient anatomy, as input to train the deep neural networks. We implement and compare two deep learning networks that predict with two different beam configuration modalities. Methods: We generated Pareto optimal plans for 70 patients with prostate cancer. We used fluence map optimization to generate 500 IMRT plans that sampled the Pareto surface for each patient, for a total of 35 000 plans. We studied and compared two different models, Models I and II. Although they both used the same anatomical structuresincluding the planning target volume (PTV), organs at risk (OARs), and bodythese models were designed with two different methods for representing beam angles. Model I directly uses beam angles as a second input to the network as a binary vector. Model II converts the beam angles into beam doses that are conformal to the PTV. We divided the 70 patients into 54 training, 6 validation, and 10 testing patients, thus yielding 27 000 training, 3000 validation, and 5000 testing plans. Mean square loss (MSE) was taken as the loss function. We used the Adam optimizer with a default learning rate of 0.01 to optimize the network's performance. We evaluated the models' performance by comparing their predicted dose distributions with the ground truth (Pareto optimal) dose distribution, in terms of dose volume histogram (DVH) plots and evaluation metrics such as PTV D 98 , D 95 , D 50 , D 2 , D max , D mean , Paddick Conformation Number, R50, and Homogeneity index. Results: Our deep learning models predicted voxel-level dose distributions that precisely matched the ground truth dose distributions. The DVHs generated also precisely matched the ground truth. Evaluation metrics such as PTV statistics, dose conformity, dose spillage (R50), and homogeneity index also confirmed the accuracy of PTV curves on the DVH. Quantitatively, Model I's prediction error of 0.043 (confirmation), 0.043 (homogeneity), 0.327 (R50), 2.80% (D95), 3.90% (D98), 0.6% (D50), and 1.10% (D2) was lower than that of Model II, which obtained 0.076 (confirmation), 0.058 (homogeneity), 0.626 (R50), 7.10% (D95), 6.50% (D98), 8.40% (D50), and 6.30% (D2). Model I also outperformed Model II in terms of the mean dose error and the max dose error on the PTV, bladder, rectum, left femoral head, and right femoral head. Conclusions: Treatment planners who use our models will be able to use deep learning to control the trade-offs between the PTV and OAR...

show abstract

“…When putting new patients' CT images and structure labels into the constructed model, the predicted dose distribution could be obtained and exported as the output, which is then further converted to yield the ultimate deliverable plans. Kajikawa et al (55) found that the dose predicted with the 3D CNN model was superior or comparable with the dose distribution generated by RapidPlan TM for prostate cancer IMRT plans using only contours in planning CT. Ma et al (56) incorporated the dose distribution from a PTVonly plan, in addition to the patient's structures contour data from planning CT in their deep CNN-based dose prediction model. The prediction results were better than the contoursbased method.…”

Section: From Dvh Prediction To Dose Distribution Prediction From Thmentioning

confidence: 99%

A Review on Application of Deep Learning Algorithms in External Beam Radiotherapy Automated Treatment Planning

et al. 2020

View full text Add to dashboard Cite

Treatment planning plays an important role in the process of radiotherapy (RT). The quality of the treatment plan directly and significantly affects patient treatment outcomes. In the past decades, technological advances in computer and software have promoted the development of RT treatment planning systems with sophisticated dose calculation and optimization algorithms. Treatment planners now have greater flexibility in designing highly complex RT treatment plans in order to mitigate the damage to healthy tissues better while maximizing radiation dose to tumor targets. Nevertheless, treatment planning is still largely a time-inefficient and labor-intensive process in current clinical practice. Artificial intelligence, including machine learning (ML) and deep learning (DL), has been recently used to automate RT treatment planning and has gained enormous attention in the RT community due to its great promises in improving treatment planning quality and efficiency. In this article, we reviewed the historical advancement, strengths, and weaknesses of various DL-based automated RT treatment planning techniques. We have also discussed the challenges, issues, and potential research directions of DL-based automated RT treatment planning techniques.

show abstract

A convolutional neural network approach for IMRT dose distribution prediction in prostate cancer patients

Cited by 46 publications

References 37 publications

Environmental Remediation of the difficult-to-return zone in Tomioka Town, Fukushima Prefecture

Environmental Remediation of the difficult-to-return zone in Tomioka Town, Fukushima Prefecture

Using deep learning to predict beam‐tunable Pareto optimal dose distribution for intensity‐modulated radiation therapy

A Review on Application of Deep Learning Algorithms in External Beam Radiotherapy Automated Treatment Planning

Contact Info

Product

Resources

About