Investigation of the optimum location of external markers for patient setup accuracy enhancement at external beam radiotherapy.

In external beam radiotherapy, one of the most common and reliable methods for patient geometrical setup and/or predicting the tumor location is use of external markers. In this study, the main challenging issue is increasing the accuracy of patient setup by investigating external markers location. Since the location of each external marker may yield different patient setup accuracy, it is important to assess different locations of external markers using appropriate selective algorithms. To do this, two commercially available algorithms entitled a) canonical correlation analysis (CCA) and b) principal component analysis (PCA) were proposed as input selection algorithms. They work on the basis of maximum correlation coefficient and minimum variance between given datasets. The proposed input selection algorithms work in combination with an adaptive neuro-fuzzy inference system (ANFIS) as a correlation model to give patient positioning information as output. Our proposed algorithms provide input file of ANFIS correlation model accurately. The required dataset for this study was prepared by means of a NURBS-based 4D XCAT anthropomorphic phantom that can model the shape and structure of complex organs in human body along with motion information of dynamic organs. Moreover, a database of four real patients undergoing radiation therapy for lung cancers was utilized in this study for validation of proposed strategy. Final analyzed results demonstrate that input selection algorithms can reasonably select specific external markers from those areas of the thorax region where root mean square error (RMSE) of ANFIS model has minimum values at that given area. It is also found that the selected marker locations lie closely in those areas where surface point motion has a large amplitude and a high correlation.

Optimum location of external markers using feature selection algorithms for real-time tumor tracking in external-beam radiotherapy: a virtual phantom study.

In external-beam radiotherapy, using external markers is one of the most reliable tools to predict tumor position, in clinical applications. The main challenge in this approach is tumor motion tracking with highest accuracy that depends heavily on external markers location, and this issue is the objective of this study. Four commercially available feature selection algorithms entitled 1) Correlation-based Feature Selection, 2) Classifier, 3) Principal Components, and 4) Relief were proposed to find optimum location of external markers in combination with two "Genetic" and "Ranker" searching procedures. The performance of these algorithms has been evaluated using four-dimensional extended cardiac-torso anthropomorphic phantom. Six tumors in lung, three tumors in liver, and 49 points on the thorax surface were taken into account to simulate internal and external motions, respectively. The root mean square error of an adaptive neuro-fuzzy inference system (ANFIS) as prediction model was considered as metric for quantitatively evaluating the performance of proposed feature selection algorithms. To do this, the thorax surface region was divided into nine smaller segments and predefined tumors motion was predicted by ANFIS using external motion data of given markers at each small segment, separately. Our comparative results showed that all feature selection algorithms can reasonably select specific external markers from those segments where the root mean square error of the ANFIS model is minimum. Moreover, the performance accuracy of proposed feature selection algorithms was compared, separately. For this, each tumor motion was predicted using motion data of those external markers selected by each feature selection algorithm. Duncan statistical test, followed by F-test, on final results reflected that all proposed feature selection algorithms have the same performance accuracy for lung tumors. But for liver tumors, a correlation-based feature selection algorithm, in combination with a genetic search algorithm, proved to yield best performance accuracy for selecting optimum markers.

Investigation the Efficacy of Fuzzy Logic Implementation at Image-Guided Radiotherapy.

At image-guided radiotherapy, technique, different imaging, and monitoring systems are utilized for (i) organs border detection and tumor delineation during the treatment planning process and (ii) patient setup and tumor localization at pretreatment step and (iii) real-time tumor motion tracking for dynamic thorax tumors during the treatment. In this study, the effect of fuzzy logic is quantitatively investigated at different steps of image-guided radiotherapy. Fuzzy logic-based models and algorithms have been implemented at three steps, and the obtained results are compared with commonly available strategies. Required data are (i) real patients treated with Synchrony Cyberknife system at Georgetown University Hospital for real-time tumor motion prediction, (ii) computed tomography images taken from real patients for geometrical setup, and also (iii) tomography images of an anthropomorphic phantom for tumor delineation process. In real-time tumor tracking, the targeting error averages of the fuzzy correlation model in comparison with the Cyberknife modeler are 4.57 mm and 8.97 mm, respectively, for a given patient that shows remarkable error reduction. In the case of patient geometrical setup, the fuzzy logic-based algorithm has better influence in comparing with the artificial neural network, while the setup error averages is reduced from 1.47 to 0.4432 mm using the fuzzy logic-based method, for a given patient.Finally, the obtained results show that the fuzzy logic based image processing algorithm exhibits much better performance for edge detection compared to four conventional operators. This study is an effort to show that fuzzy logic based algorithms are also highly applicable at image-guided radiotherapy as one of the important treatment modalities for tumor delineation, patient setup error reduction, and intrafractional motion error compensation due to their inherent properties.

Use of machine learning methods for prediction of acute toxicity in organs at risk following prostate radiotherapy.

PURPOSE: The goal of this study is to investigate the advantages of large scale optimization methods vs conventional classification techniques in predicting acute toxicity for urinary bladder and rectum due to prostate irradiation. METHODS: Clinical and dosimetric data of 321 patients undergoing prostate conformal radiotherapy were recorded. Gastro-intestinal and genito-urinary acute toxicities were scored according to the Radiation Therapy Oncology Group/European Organization for Research and Treatment of Cancer (RTOG/EORTC) scale. Patients were classified in two categories to separate mild (Grade < 2) from severe toxicity levels (Grade > 2). Machine learning methods at different complexity were implemented to predict toxicity as a function of multiple variables. The first approach consisted of a large scale optimization method, based on genetic algorithms (GAs) and artificial neural networks (ANN). The second approach was a binary classifier based on support vector machines (SVM). RESULTS: The ANN and SVM-based solutions showed comparable prediction accuracy, exhibiting an area under the receiver operating characteristic (ROC) curve of 0.7. Different sensitivity and specificity features were measured for the two approaches. The ANN algorithm showed enhanced sensitivity if combined with appropriate classification criteria. CONCLUSIONS: The results demonstrate that high sensitivity in toxicity prediction can be achieved with optimized ANNs, that are put forward to represent a valuable support in medical decisions. Future studies will be focused on enlarging the available patient database to increase the reliability of toxicity prediction algorithms and to define optimal classification criteria.

A Feasibility Study on Ribs as Anatomical Landmarks for Motion Tracking of Lung and Liver Tumors at External Beam Radiotherapy.

At external beam radiotherapy for some tumors located at thorax region due to lack of information in gray scale fluoroscopic images tumor position determination is problematic. One of the clinical strategies is to implant clip as internal fiducial marker inside or near tumor to represent tumor position while the contrast of implanted clip is highly observable rather than tumor. As alternative, using natural anatomical landmarks located at thorax region of patient body is proposed to extract tumor position information without implanting clips that is invasive method with possible side effect. Among natural landmarks, ribs of rib-cage structure that result proper visualization at X-ray images may be optimal as representative for tumor motion. In this study, we investigated the existence of possible correlation between ribs as natural anatomical landmarks and various lung and liver tumors located at different sites as challenging issue. A simulation study was performed using data extracted from 4-dimensional extended cardiac-torso anthropomorphic phantom that is able to simulate motion effect of dynamic organs, as well. Several tumor sites with predefined distances originated from chosen ribs at anterior-posterior direction were simulated at 3 upper, middle, and lower parts of chest. Correlation coefficient between ribs and tumors was calculated to investigate the robustness of ribs as anatomical landmarks for tumor motion tracking. Moreover, a consistent correlation model was taken into account to track tumor motion with a rib as best candidate among selected ribs. Final results represent availability of using rib cage as anatomical landmark to track lung and liver tumors in a noninvasive way. Observations of our calculations showed a proper correlation between tumors and ribs while the degree of this correlation is changing depends on tumor site while lung tumors are more varied and complex with less correlation with ribs motion against liver tumors.

Development of a Synthetic Adaptive Neuro-Fuzzy Prediction Model for Tumor Motion Tracking in External Radiotherapy by Evaluating Various Data Clustering Algorithms.

In image guided radiotherapy, in order to reach a prescribed uniform dose in dynamic tumors at thorax region while minimizing the amount of additional dose received by the surrounding healthy tissues, tumor motion must be tracked in real-time. Several correlation models have been proposed in recent years to provide tumor position information as a function of time in radiotherapy with external surrogates. However, developing an accurate correlation model is still a challenge. In this study, we proposed an adaptive neuro-fuzzy based correlation model that employs several data clustering algorithms for antecedent parameters construction to avoid over-fitting and to achieve an appropriate performance in tumor motion tracking compared with the conventional models. To begin, a comparative assessment is done between seven nuero-fuzzy correlation models each constructed using a unique data clustering algorithm. Then, each of the constructed models are combined within an adaptive sevenfold synthetic model since our tumor motion database has high degrees of variability and that each model has its intrinsic properties at motion tracking. In the proposed sevenfold synthetic model, best model is selected adaptively at pre-treatment. The model also updates the steps for each patient using an automatic model selectivity subroutine. We tested the efficacy of the proposed synthetic model on twenty patients (divided equally into two control and worst groups) treated with CyberKnife synchrony system. Compared to Cyberknife model, the proposed synthetic model resulted in 61.2% and 49.3% reduction in tumor tracking error in worst and control group, respectively. These results suggest that the proposed model selection program in our synthetic neuro-fuzzy model can significantly reduce tumor tracking errors. Numerical assessments confirmed that the proposed synthetic model is able to track tumor motion in real time with high accuracy during treatment.

Investigation of the robustness of adaptive neuro-fuzzy inference system for tracking moving tumors in external radiotherapy.

In external radiotherapy of dynamic targets such as lung and breast cancers, accurate correlation models are utilized to extract real time tumor position by means of external surrogates in correlation with the internal motion of tumors. In this study, a correlation method based on the neuro-fuzzy model is proposed to correlate the input external motion data with internal tumor motion estimation in real-time mode, due to its robustness in motion tracking. An initial test of the performance of this model was reported in our previous studies. In this work by implementing some modifications it is resulted that ANFIS is still robust to track tumor motion more reliably by reducing the motion estimation error remarkably. After configuring new version of our ANFIS model, its performance was retrospectively tested over ten patients treated with Synchrony Cyberknife system. In order to assess the performance of our model, the predicted tumor motion as model output was compared with respect to the state of the art model. Final analyzed results show that our adaptive neuro-fuzzy model can reduce tumor tracking errors more significantly, as compared with ground truth database and even tumor tracking methods presented in our previous works.