Modeling plan-related clinical complications using machine learning tools in a multiplan IMRT framework

Embedding machine learning based toxicity models within radiotherapy treatment plan optimization

Objective.This study addresses radiation-induced toxicity (RIT) challenges in radiotherapy (RT) by developing a personalized treatment planning framework. It leverages patient-specific data and dosimetric information to create an optimization model that limits adverse side effects using constraints learned from historical data.Approach.The study uses the optimization with constraint learning (OCL) framework, incorporating patient-specific factors into the optimization process. It consists of three steps: optimizing the baseline treatment plan using population-wide dosimetric constraints; training a machine learning (ML) model to estimate the patient's RIT for the baseline plan; and adapting the treatment plan to minimize RIT using ML-learned patient-specific constraints. Various predictive models, including classification trees, ensembles of trees, and neural networks, are applied to predict the probability of grade 2+ radiation pneumonitis (RP2+) for non-small cell lung (NSCLC) cancer patients three months post-RT. The methodology is assessed with four high RP2+ risk NSCLC patients, with the goal of optimizing the dose distribution to constrain the RP2+ outcome below a pre-specified threshold. Conventional and OCL-enhanced plans are compared based on dosimetric parameters and predicted RP2+ risk. Sensitivity analysis on risk thresholds and data uncertainty is performed using a toy NSCLC case.Main results.Experiments show the methodology's capacity to directly incorporate all predictive models into RT treatment planning. In the four patients studied, mean lung dose and V20 were reduced by an average of 1.78 Gy and 3.66%, resulting in an average RP2+ risk reduction from 95% to 42%. Notably, this reduction maintains tumor coverage, although in two cases, sparing the lung slightly increased spinal cord max-dose (0.23 and 0.79 Gy).Significance.By integrating patient-specific information into learned constraints, the study significantly reduces adverse side effects like RP2+ without compromising target coverage. This unified framework bridges the gap between predicting toxicities and optimizing treatment plans in personalized RT decision-making.

Improvement of IMRT QA prediction using imaging-based neural architecture search

PURPOSE

Machine learning has been used to predict the gamma passing rate of Intensity-modulated radiation therapy (IMRT) QA results. In this work, we applied a novel neural architecture search to automatically tune and search for the best deep neural networks instead of using hand-designed deep learning architectures.

METHOD AND MATERIALS

One hundred and eighty-two IMRT plans were created and delivered with portal dosimetry. A total of 1497 fields for multiple treatment sites were delivered and measured by portal imagers. Gamma criteria of 2%/2mm with a 5% threshold were used. Fluence maps calculated for each plan were used as inputs to a convolution neural network (CNN). Auto-Keras was implemented to search for the best CNN architecture for fluence image regression. The network morphism was adopted in the searching process, in which the base models were ResNet and DenseNet. The performance of this CNN approach was compared with tree-based machine learning models previously developed for this application, using the same data set.

RESULTS

The deep-learning-based approach had 98.3% of predictions within 3% of the measured 2%/2mm gamma passing rates with a maximum error of 3.1% and a mean absolute error of less than 1%. Our results show that this novel architecture search approach achieves comparable performance to the machine-learning-based approaches with handcrafted features.

CONCLUSIONS

We implemented a novel CNN model using imaging-based neural architecture for IMRT QA prediction. The imaging-based deep-learning method does not require manual extraction of relevant features and is able to automatically select the best network architecture. This article is protected by copyright. All rights reserved.

Personalized brachytherapy dose reconstruction using deep learning

BACKGROUND AND PURPOSE

Accurate calculation of the absorbed dose delivered to the tumor and normal tissues improves treatment gain factor, which is the major advantage of brachytherapy over external radiation therapy. To address the simplifications of TG-43 assumptions that ignore the dosimetric impact of medium heterogeneities, we proposed a deep learning (DL)-based approach, which improves the accuracy while requiring a reasonable computation time.

MATERIALS AND METHODS

We developed a Monte Carlo (MC)-based personalized brachytherapy dosimetry simulator (PBrDoseSim), deployed to generate patient-specific dose distributions. A deep neural network (DNN) was trained to predict personalized dose distributions derived from MC simulations, serving as ground truth. The paired channel input used for the training is composed of dose distribution kernel in water medium along with the full-volumetric density maps obtained from CT images reflecting medium heterogeneity.

RESULTS

The predicted single-dwell dose kernels were in good agreement with MC-based kernels serving as reference, achieving a mean relative absolute error (MRAE) and mean absolute error (MAE) of 1.16 ± 0.42% and 4.2 ± 2.7 × 10^-4 (Gy.sec^-1/voxel), respectively. The MRAE of the dose volume histograms (DVHs) between the DNN and MC calculations in the clinical target volume were 1.8 ± 0.86%, 0.56 ± 0.56%, and 1.48 ± 0.72% for D90, V150, and V100, respectively. For bladder, sigmoid, and rectum, the MRAE of D5cc between the DNN and MC calculations were 2.7 ± 1.7%, 1.9 ± 1.3%, and 2.1 ± 1.7%, respectively.

CONCLUSION

The proposed DNN-based personalized brachytherapy dosimetry approach exhibited comparable performance to the MC method while overcoming the computational burden of MC calculations and oversimplifications of TG-43.

Machine Learning-Based Models for Prediction of Toxicity Outcomes in Radiotherapy

In order to limit radiotherapy (RT)-related side effects, effective toxicity prediction and assessment schemes are essential. In recent years, the growing interest toward artificial intelligence and machine learning (ML) within the science community has led to the implementation of innovative tools in RT. Several researchers have demonstrated the high performance of ML-based models in predicting toxicity, but the application of these approaches in clinics is still lagging, partly due to their low interpretability. Therefore, an overview of contemporary research is needed in order to familiarize practitioners with common methods and strategies. Here, we present a review of ML-based models for predicting and classifying RT-induced complications from both a methodological and a clinical standpoint, focusing on the type of features considered, the ML methods used, and the main results achieved. Our work overviews published research in multiple cancer sites, including brain, breast, esophagus, gynecological, head and neck, liver, lung, and prostate cancers. The aim is to define the current state of the art and main achievements within the field for both researchers and clinicians.

Deep Learning: A Review for the Radiation Oncologist

Introduction: Deep Learning (DL) is a machine learning technique that uses deep neural networks to create a model. The application areas of deep learning in radiation oncology include image segmentation and detection, image phenotyping, and radiomic signature discovery, clinical outcome prediction, image dose quantification, dose-response modeling, radiation adaptation, and image generation. In this review, we explain the methods used in DL and perform a literature review using the Medline database to identify studies using deep learning in radiation oncology. The search was conducted in April 2018, and identified studies published between 1997 and 2018, strongly skewed toward 2015 and later. Methods: A literature review was performed using PubMed/Medline in order to identify important recent publications to be synthesized into a review of the current status of Deep Learning in radiation oncology, directed at a clinically-oriented reader. The search strategy included the search terms "radiotherapy" and "deep learning." In addition, reference lists of selected articles were hand searched for further potential hits of relevance to this review. The search was conducted in April 2018, and identified studies published between 1997 and 2018, strongly skewed toward 2015 and later. Results: Studies using DL for image segmentation were identified in Brain (n = 2), Head and Neck (n = 3), Lung (n = 6), Abdominal (n = 2), and Pelvic (n = 6) cancers. Use of Deep Learning has also been reported for outcome prediction, such as toxicity modeling (n = 3), treatment response and survival (n = 2), or treatment planning (n = 5). Conclusion: Over the past few years, there has been a significant number of studies assessing the performance of DL techniques in radiation oncology. They demonstrate how DL-based systems can aid clinicians in their daily work, be it by reducing the time required for or the variability in segmentation, or by helping to predict treatment outcomes and toxicities. It still remains to be seen when these techniques will be employed in routine clinical practice.

Modern radiotherapy for head and neck cancer

Radiation therapy (RT) plays a key role in curative-intent treatments for head and neck cancers. Its use is indicated as a sole therapy in early stage tumors or in combination with surgery or concurrent chemotherapy in advanced stages. Recent technologic advances have resulted in both improved oncologic results and expansion of the indications for RT in clinical practice. Despite this, RT administered to the head and neck region is still burdened by a high rate of acute and late side effects. Moreover, about 50% of patients with high-risk disease experience loco-regional recurrence within 3 years of follow-up. Therefore, in recent decades, efforts have been dedicated to optimize the cost/benefit ratio of RT in this subset of patients. The aim of the present review was to highlight modern concepts of RT for head and neck cancers considering both the technological advances that have been achieved and recent knowledge that has informed the biological interaction between radiation and both tumor and healthy tissues.

Radiomics and Machine Learning for Radiotherapy in Head and Neck Cancers

Introduction: An increasing number of parameters can be considered when making decisions in oncology. Tumor characteristics can also be extracted from imaging through the use of radiomics and add to this wealth of clinical data. Machine learning can encompass these parameters and thus enhance clinical decision as well as radiotherapy workflow. Methods: We performed a description of machine learning applications at each step of treatment by radiotherapy in head and neck cancers. We then performed a systematic review on radiomics and machine learning outcome prediction models in head and neck cancers. Results: Machine Learning has several promising applications in treatment planning with automatic organ at risk delineation improvements and adaptative radiotherapy workflow automation. It may also provide new approaches for Normal Tissue Complication Probability models. Radiomics may provide additional data on tumors for improved machine learning powered predictive models, not only on survival, but also on risk of distant metastasis, in field recurrence, HPV status and extra nodal spread. However, most studies provide preliminary data requiring further validation. Conclusion: Promising perspectives arise from machine learning applications and radiomics based models, yet further data are necessary for their implementation in daily care.

Deep nets vs expert designed features in medical physics: An IMRT QA case study

PURPOSE

The purpose of this study was to compare the performance of Deep Neural Networks against a technique designed by domain experts in the prediction of gamma passing rates for Intensity Modulated Radiation Therapy Quality Assurance (IMRT QA).

METHOD

A total of 498 IMRT plans across all treatment sites were planned in Eclipse version 11 and delivered using a dynamic sliding window technique on Clinac iX or TrueBeam Linacs. Measurements were performed using a commercial 2D diode array, and passing rates for 3%/3 mm local dose/distance-to-agreement (DTA) were recorded. Separately, fluence maps calculated for each plan were used as inputs to a convolution neural network (CNN). The CNNs were trained to predict IMRT QA gamma passing rates using TensorFlow and Keras. A set of model architectures, inspired by the convolutional blocks of the VGG-16 ImageNet model, were constructed and implemented. Synthetic data, created by rotating and translating the fluence maps during training, was created to boost the performance of the CNNs. Dropout, batch normalization, and data augmentation were utilized to help train the model. The performance of the CNNs was compared to a generalized Poisson regression model, previously developed for this application, which used 78 expert designed features.

RESULTS

Deep Neural Networks without domain knowledge achieved comparable performance to a baseline system designed by domain experts in the prediction of 3%/3 mm Local gamma passing rates. An ensemble of neural nets resulted in a mean absolute error (MAE) of 0.70 ± 0.05 and the domain expert model resulted in a 0.74 ± 0.06.

CONCLUSIONS

Convolutional neural networks (CNNs) with transfer learning can predict IMRT QA passing rates by automatically designing features from the fluence maps without human expert supervision. Predictions from CNNs are comparable to a system carefully designed by physicist experts.

IMRT QA using machine learning: A multi-institutional validation

Purpose

To validate a machine learning approach to Virtual intensity‐modulated radiation therapy (IMRT) quality assurance (QA) for accurately predicting gamma passing rates using different measurement approaches at different institutions.

Methods

A Virtual IMRT QA framework was previously developed using a machine learning algorithm based on 498 IMRT plans, in which QA measurements were performed using diode‐array detectors and a 3%local/3 mm with 10% threshold at Institution 1. An independent set of 139 IMRT measurements from a different institution, Institution 2, with QA data based on portal dosimetry using the same gamma index, was used to test the mathematical framework. Only pixels with ≥10% of the maximum calibrated units (CU) or dose were included in the comparison. Plans were characterized by 90 different complexity metrics. A weighted poison regression with Lasso regularization was trained to predict passing rates using the complexity metrics as input.

Results

The methodology predicted passing rates within 3% accuracy for all composite plans measured using diode‐array detectors at Institution 1, and within 3.5% for 120 of 139 plans using portal dosimetry measurements performed on a per‐beam basis at Institution 2. The remaining measurements (19) had large areas of low CU, where portal dosimetry has a larger disagreement with the calculated dose and as such, the failure was expected. These beams need further modeling in the treatment planning system to correct the under‐response in low‐dose regions. Important features selected by Lasso to predict gamma passing rates were as follows: complete irradiated area outline (CIAO), jaw position, fraction of MLC leafs with gaps smaller than 20 or 5 mm, fraction of area receiving less than 50% of the total CU, fraction of the area receiving dose from penumbra, weighted average irregularity factor, and duty cycle.

Conclusions

We have demonstrated that Virtual IMRT QA can predict passing rates using different measurement techniques and across multiple institutions. Prediction of QA passing rates can have profound implications on the current IMRT process.

Predictive modelling analysis for development of a radiotherapy decision support system in prostate cancer: a preliminary study

Purpose

The aim of this study is to develop predictive models to predict organ at risk (OAR) complication level, classification of OAR dose-volume and combination of this function with our in-house developed treatment decision support system.

Materials and methods

We analysed the support vector machine and decision tree algorithm for predicting OAR complication level and toxicity in order to integrate this function into our in-house radiation treatment planning decision support system. A total of 12 TomoTherapyTM treatment plans for prostate cancer were established, and a hundred modelled plans were generated to analyse the toxicity prediction for bladder and rectum.

Results

The toxicity prediction algorithm analysis showed 91·0% accuracy in the training process. A scatter plot for bladder and rectum was obtained by 100 modelled plans and classification result derived. OAR complication level was analysed and risk factor for 25% bladder and 50% rectum was detected by decision tree. Therefore, it was shown that complication prediction of patients using big data-based clinical information is possible.

Conclusion

We verified the accuracy of the tested algorithm using prostate cancer cases. Side effects can be minimised by applying this predictive modelling algorithm with the planning decision support system for patient-specific radiotherapy planning.

Confidence limits for patient-specific IMRT dose QA: a multi-institutional study in Korea

This study aims to investigate tolerance levels for patient-specific IMRT dose QA (DQA) using the confidence limits (CL) determined by a multi-institutional study. Eleven institutions participated in the multi-institutional study in Korea. A total of 155 DQA measurements, consisting of point-dose differences (high- and low-dose regions) and gamma passing rates (composite and per-field) for IMRT patients with brain, head and neck (H&N), abdomen, and prostate cancers were examined. The Shapiro-Wilk test was used to evaluate the normality of data grouped by the treatment sites and the DQA methods. The confidence limit coefficients in cases of the normal distribution, and the two-sided Student's t-distribution were applied to determine the confidence limits for the grouped data. The Spearman's test was applied to assess the sensitivity of DQA results within the limited groups. The differences in CLs between the two confidence coefficients based on the normal and t-distributions were negligible for the point-dose data and the gamma passing rates with 3%/3 mm criteria. However, with 2%/2 mm criteria, the difference in CLs were 1.6% and 2.2% for composite and per-field measurements, respectively. This resulted from the large standard deviation and the more sensitive criteria of 2%/2 mm. There was no noticeable correlation among the different QA methods. Our multi-institutional study suggested that the CL was not a suitable metric for defining the tolerance level when the statistics of the sample group did not follow the normality and had a large standard deviation.

Automatic assessment of average diaphragm motion trajectory from 4DCT images through machine learning

To automatically estimate average diaphragm motion trajectory (ADMT) based on four-dimensional computed tomography (4DCT), facilitating clinical assessment of respiratory motion and motion variation and retrospective motion study. We have developed an effective motion extraction approach and a machine-learning-based algorithm to estimate the ADMT. Eleven patients with 22 sets of 4DCT images (4DCT1 at simulation and 4DCT2 at treatment) were studied. After automatically segmenting the lungs, the differential volume-per-slice (dVPS) curves of the left and right lungs were calculated as a function of slice number for each phase with respective to the full-exhalation. After 5-slice moving average was performed, the discrete cosine transform (DCT) was applied to analyze the dVPS curves in frequency domain. The dimensionality of the spectrum data was reduced by using several lowest frequency coefficients (f_v) to account for most of the spectrum energy (Σf_v²). Multiple linear regression (MLR) method was then applied to determine the weights of these frequencies by fitting the ground truth-the measured ADMT, which are represented by three pivot points of the diaphragm on each side. The 'leave-one-out' cross validation method was employed to analyze the statistical performance of the prediction results in three image sets: 4DCT1, 4DCT2, and 4DCT1 + 4DCT2. Seven lowest frequencies in DCT domain were found to be sufficient to approximate the patient dVPS curves (R = 91%-96% in MLR fitting). The mean error in the predicted ADMT using leave-one-out method was 0.3 ± 1.9 mm for the left-side diaphragm and 0.0 ± 1.4 mm for the right-side diaphragm. The prediction error is lower in 4DCT2 than 4DCT1, and is the lowest in 4DCT1 and 4DCT2 combined. This frequency-analysis-based machine learning technique was employed to predict the ADMT automatically with an acceptable error (0.2 ± 1.6 mm). This volumetric approach is not affected by the presence of the lung tumors, providing an automatic robust tool to evaluate diaphragm motion.

A support vector machine tool for adaptive tomotherapy treatments: Prediction of head and neck patients criticalities

PURPOSE

Adaptive radiation therapy (ART) is an advanced field of radiation oncology. Image-guided radiation therapy (IGRT) methods can support daily setup and assess anatomical variations during therapy, which could prevent incorrect dose distribution and unexpected toxicities. A re-planning to correct these anatomical variations should be done daily/weekly, but to be applicable to a large number of patients, still require time consumption and resources. Using unsupervised machine learning on retrospective data, we have developed a predictive network, to identify patients that would benefit of a re-planning.

METHODS

1200 MVCT of 40 head and neck (H&N) cases were re-contoured, automatically, using deformable hybrid registration and structures mapping. Deformable algorithm and MATLAB(®) homemade machine learning process, developed, allow prediction of criticalities for Tomotherapy treatments.

RESULTS

Using retrospective analysis of H&N treatments, we have investigated and predicted tumor shrinkage and organ at risk (OAR) deformations. Support vector machine (SVM) and cluster analysis have identified cases or treatment sessions with potential criticalities, based on dose and volume discrepancies between fractions. During 1st weeks of treatment, 84% of patients shown an output comparable to average standard radiation treatment behavior. Starting from the 4th week, significant morpho-dosimetric changes affect 77% of patients, suggesting need for re-planning. The comparison of treatment delivered and ART simulation was carried out with receiver operating characteristic (ROC) curves, showing monotonous increase of ROC area.

CONCLUSIONS

Warping methods, supported by daily image analysis and predictive tools, can improve personalization and monitoring of each treatment, thereby minimizing anatomic and dosimetric divergences from initial constraints.

The minimum knowledge base for predicting organ-at-risk dose-volume levels and plan-related complications in IMRT planning

IMRT treatment planning requires consideration of two competing objectives: achieving the required amount of radiation for the planning target volume and minimizing the amount of radiation delivered to all other tissues. It is important for planners to understand the tradeoff between competing factors so that the time-consuming human interaction loop (plan-evaluate-modify) can be eliminated. Treatment-plan-surface models have been proposed as a decision support tool to aid treatment planners and clinicians in choosing between rival treatment plans in a multi-plan environment. In this paper, an empirical approach is introduced to determine the minimum number of treatment plans (minimum knowledge base) required to build accurate representations of the IMRT plan surface in order to predict organ-at-risk (OAR) dose-volume (DV) levels and complications as a function of input DV constraint settings corresponding to all involved OARs in the plan. We have tested our approach on five head and neck patients and five whole pelvis/prostate patients. Our results suggest that approximately 30 plans were sufficient to predict DV levels with less than 3% relative error in both head and neck and whole pelvis/prostate cases. In addition, approximately 30-60 plans were sufficient to predict saliva flow rate with less than 2% relative error and to classify rectal bleeding with an accuracy of 90%.