- Email: [email protected]
Dosimetric comparison of intensity-modulated radiotherapy versus 3D conformal radiotherapy in patients with head-and-neck cancer
Physica Medica 31 (2015) S1–S14
Contents lists available at ScienceDirect
Physica Medica j o u r n a l h o m e p a g e : h t t p : / / w w w. p h y s i c a m e d i c a . c o m
SAAPMB 2015 Oral Presentations O.1 MASS-SPECIFIC THRESHOLD VALUES OF GLOBAL MINIMA FOR CONVEX ENERGY FUNCTIONALS WITH AN INTERACTIVE BREAST MASS SEGMENTATION MODEL S.N. Acho *, W.I.D. Rae. Department of Medical Physics, University of the Free State, Bloemfontein, South Africa Introduction: A convex active contour model requires a predefined threshold value for the global minimizer of the energy functional. In mass segmentation, a fixed threshold value for the global minima of a database of mass lesions is not feasible. Furthermore, hand tuning each threshold value for optimum mass boundary delineation is time consuming. The proposed method targets the highlighted focal region of each mass lesion by means of its probability matrix, P (x, y), to provide a mass-specific threshold value of the global minimum for its convex energy functional. We qualitatively compare our results with the Chan–Vese segmentation method and a published global segmentation model on masses found on direct digital mammograms. Materials and Methods: The regional term of the convex active contour model maximizes the posterior partitioning probability for binary segmentation. It has a binary classifier which propagates the contour whenever the statistical models of the foreground and background are similar. Suppose P (x, y) is binary thresholded with the particle swarm optimization method to obtain a value, T1; then, we define the optimal threshold value for the global minimizer as the mean intensity value of all the pixels whose probability values are greater than T1. Results: The mean Jaccard similarity indices were 0.89 ± 0.07 for the proposed/Chan–Vese method and 0.88 ± 0.06 for the proposed/published segmentation model. The mean Euclidean distance between the Fourier descriptors of the segmented areas was 0.05 ± 0.03 for the proposed/Chan– Vese method and 0.06 ± 0.04 for the proposed/published segmentation model. Conclusion: This method is efficient and avoids the problems associated with the placement of the initial level set contours and contour reinitialization. Moreover optimum segmentation results are realized for all masses as opposed to the fixed threshold value of 0.5, proposed by other researchers. Keywords: Segmentation, Chan–Vese, Mass, Convex active contour O.2 EVOLUTION OF THE IDEAL WORKFLOW: A CASE STUDY OF PROSTATE CANCER RADIOTHERAPY AT SGMC R. Agbeve *, C.O. Doudoo, G.F. Acquah, P. Oppong, A. Shulman. Sweden Ghana Medical Centre (SGMC) Ltd, Tema Municipal, Ghana Introduction: Establishing a new oncology centre involves the structuring of an effective radiotherapy department, which depends on welldefined protocols. Using prostate cancer radiotherapy at SGMC Ltd as a template, this paper aims to portray the concept of a ‘gold-standard’ workflow as one which enhances patient satisfaction, is stable and safe and above all is developed to suit individual hospitals’ strengths. Materials and Methods: Five key areas of the entire prostate cancer radiotherapy process were examined to highlight problem areas in the existing
* Corresponding author.
protocols, i.e. CT patient preparation; PTV and OAR delineation using CT versus CT/MRI image fusion; plan optimization and dealing with low dose areas within the CTV caused by gold marker artefacts; incorporation of portal imaging monitor units (MUs) into treatment MUs and enhancement of the quality of portal images (on the iView GT system) during treatment delivery. Results: The continual review of existing protocols in the above areas has had a profound effect on the quality of prostate radiotherapy at SGMC Ltd. An example is MRI/CT fusion which was proven to enable better PTV and OAR delineation than CT images alone. Overall, the study led to a continual positive evolution of our workflow, eventually culminating in the adoption of the current work protocols. Conclusion: An ideal workflow is one which anticipates and addresses problems arising as the patient(s)/data proceeds through the imaging, dosimetry and treatment cycle. Indeed, there is no purely superior or inferior workflow because a radiotherapy centre should adopt the most suitable workflow, dictated by its resources. Keywords: Radiotherapy, Workflow, CT, MRI, Fusion O.3 DOSIMETRIC COMPARISON OF INTENSITY-MODULATED RADIOTHERAPY VERSUS 3D CONFORMAL RADIOTHERAPY IN PATIENTS WITH HEAD-ANDNECK CANCER D.M. AL Zayat *,a, E.M. Attalla b, H.S. Abouelenein c, S. Fadel d, W. Khalil e. a Department of Medical Physics, Ayadi-Al Mostakbl Oncology Center, Alex, Egypt; b Department of Medical Physics, National Cancer Institute, Cairo University, Egypt; c Department of Medical Physics, Children’s Cancer Hospital, Egypt; d Department of Oncology, Medicine Faculty, Alexandria University, Egypt; e Department of Biophysics, Faculty of Science, Cairo University, Egypt Introduction: Head and neck (H&N) cancer originates from the mucous lining of respiratory- and digestive tracts, salivary glands, and lymph nodes. H&N cancer is histologically heterogeneous and organs at risk have less tolerance to radiation. Treatment planning for advanced H&N cancer is problematic due to their complex volume shape which commonly is irregular and sometimes concave. There is also an associated need to spare critical organs like the mandible, parotid glands, brain stem and spinal cord. Nasopharyngeal carcinoma (NPC) is primarily treated with radiation therapy (RT). The basic principle of this treatment is to deliver a curative dose to the tumor while minimizing dose to surrounding structures. IMRT is a highly conformal treatment modality that is used when conventional methods of radiotherapy cannot deliver a tumor dose without exceeding critical structure tolerance. Materials and Methods: 3DCRT and IMRT plans were calculated on sixteen patients with NPC. The prescribed dose covered at least 95% of the clinical target volume (CTV). The dose prescription for nasopharyngeal carcinoma was 70 Gy to the CTV70, 63 Gy to the CTV63 and 56 Gy to the CTV56, over 35 fractions. The dose constraints to the organs at risk (OARs) were set as follows: The maximum dose to the brain stem, optic nerve and optic chiasm was 54 Gy, and the maximum spinal cord dose was 45 Gy. The mean parotid dose was 26 Gy. Results: IMRT has better coverage than 3DCRT; the doses to the spinal cord, brain stem and parotid gland from IMRT were lower than 3DCRT especially for parotid, where the dose reduction was very significant. Conclusion: Simultaneous integrated boost IMRT achieved comparable plans to 3D-CRT in complex nasopharyngeal carcinoma, IMRT treatment results in improved dose distribution within CTV compared to 3DCRT. Keywords: Nasopharyngeal carcinoma, 3DCRT, IMRT, Integrated boost
Contents lists available at ScienceDirect
Physica Medica j o u r n a l h o m e p a g e : h t t p : / / w w w. p h y s i c a m e d i c a . c o m
SAAPMB 2015 Oral Presentations O.1 MASS-SPECIFIC THRESHOLD VALUES OF GLOBAL MINIMA FOR CONVEX ENERGY FUNCTIONALS WITH AN INTERACTIVE BREAST MASS SEGMENTATION MODEL S.N. Acho *, W.I.D. Rae. Department of Medical Physics, University of the Free State, Bloemfontein, South Africa Introduction: A convex active contour model requires a predefined threshold value for the global minimizer of the energy functional. In mass segmentation, a fixed threshold value for the global minima of a database of mass lesions is not feasible. Furthermore, hand tuning each threshold value for optimum mass boundary delineation is time consuming. The proposed method targets the highlighted focal region of each mass lesion by means of its probability matrix, P (x, y), to provide a mass-specific threshold value of the global minimum for its convex energy functional. We qualitatively compare our results with the Chan–Vese segmentation method and a published global segmentation model on masses found on direct digital mammograms. Materials and Methods: The regional term of the convex active contour model maximizes the posterior partitioning probability for binary segmentation. It has a binary classifier which propagates the contour whenever the statistical models of the foreground and background are similar. Suppose P (x, y) is binary thresholded with the particle swarm optimization method to obtain a value, T1; then, we define the optimal threshold value for the global minimizer as the mean intensity value of all the pixels whose probability values are greater than T1. Results: The mean Jaccard similarity indices were 0.89 ± 0.07 for the proposed/Chan–Vese method and 0.88 ± 0.06 for the proposed/published segmentation model. The mean Euclidean distance between the Fourier descriptors of the segmented areas was 0.05 ± 0.03 for the proposed/Chan– Vese method and 0.06 ± 0.04 for the proposed/published segmentation model. Conclusion: This method is efficient and avoids the problems associated with the placement of the initial level set contours and contour reinitialization. Moreover optimum segmentation results are realized for all masses as opposed to the fixed threshold value of 0.5, proposed by other researchers. Keywords: Segmentation, Chan–Vese, Mass, Convex active contour O.2 EVOLUTION OF THE IDEAL WORKFLOW: A CASE STUDY OF PROSTATE CANCER RADIOTHERAPY AT SGMC R. Agbeve *, C.O. Doudoo, G.F. Acquah, P. Oppong, A. Shulman. Sweden Ghana Medical Centre (SGMC) Ltd, Tema Municipal, Ghana Introduction: Establishing a new oncology centre involves the structuring of an effective radiotherapy department, which depends on welldefined protocols. Using prostate cancer radiotherapy at SGMC Ltd as a template, this paper aims to portray the concept of a ‘gold-standard’ workflow as one which enhances patient satisfaction, is stable and safe and above all is developed to suit individual hospitals’ strengths. Materials and Methods: Five key areas of the entire prostate cancer radiotherapy process were examined to highlight problem areas in the existing
* Corresponding author.
protocols, i.e. CT patient preparation; PTV and OAR delineation using CT versus CT/MRI image fusion; plan optimization and dealing with low dose areas within the CTV caused by gold marker artefacts; incorporation of portal imaging monitor units (MUs) into treatment MUs and enhancement of the quality of portal images (on the iView GT system) during treatment delivery. Results: The continual review of existing protocols in the above areas has had a profound effect on the quality of prostate radiotherapy at SGMC Ltd. An example is MRI/CT fusion which was proven to enable better PTV and OAR delineation than CT images alone. Overall, the study led to a continual positive evolution of our workflow, eventually culminating in the adoption of the current work protocols. Conclusion: An ideal workflow is one which anticipates and addresses problems arising as the patient(s)/data proceeds through the imaging, dosimetry and treatment cycle. Indeed, there is no purely superior or inferior workflow because a radiotherapy centre should adopt the most suitable workflow, dictated by its resources. Keywords: Radiotherapy, Workflow, CT, MRI, Fusion O.3 DOSIMETRIC COMPARISON OF INTENSITY-MODULATED RADIOTHERAPY VERSUS 3D CONFORMAL RADIOTHERAPY IN PATIENTS WITH HEAD-ANDNECK CANCER D.M. AL Zayat *,a, E.M. Attalla b, H.S. Abouelenein c, S. Fadel d, W. Khalil e. a Department of Medical Physics, Ayadi-Al Mostakbl Oncology Center, Alex, Egypt; b Department of Medical Physics, National Cancer Institute, Cairo University, Egypt; c Department of Medical Physics, Children’s Cancer Hospital, Egypt; d Department of Oncology, Medicine Faculty, Alexandria University, Egypt; e Department of Biophysics, Faculty of Science, Cairo University, Egypt Introduction: Head and neck (H&N) cancer originates from the mucous lining of respiratory- and digestive tracts, salivary glands, and lymph nodes. H&N cancer is histologically heterogeneous and organs at risk have less tolerance to radiation. Treatment planning for advanced H&N cancer is problematic due to their complex volume shape which commonly is irregular and sometimes concave. There is also an associated need to spare critical organs like the mandible, parotid glands, brain stem and spinal cord. Nasopharyngeal carcinoma (NPC) is primarily treated with radiation therapy (RT). The basic principle of this treatment is to deliver a curative dose to the tumor while minimizing dose to surrounding structures. IMRT is a highly conformal treatment modality that is used when conventional methods of radiotherapy cannot deliver a tumor dose without exceeding critical structure tolerance. Materials and Methods: 3DCRT and IMRT plans were calculated on sixteen patients with NPC. The prescribed dose covered at least 95% of the clinical target volume (CTV). The dose prescription for nasopharyngeal carcinoma was 70 Gy to the CTV70, 63 Gy to the CTV63 and 56 Gy to the CTV56, over 35 fractions. The dose constraints to the organs at risk (OARs) were set as follows: The maximum dose to the brain stem, optic nerve and optic chiasm was 54 Gy, and the maximum spinal cord dose was 45 Gy. The mean parotid dose was 26 Gy. Results: IMRT has better coverage than 3DCRT; the doses to the spinal cord, brain stem and parotid gland from IMRT were lower than 3DCRT especially for parotid, where the dose reduction was very significant. Conclusion: Simultaneous integrated boost IMRT achieved comparable plans to 3D-CRT in complex nasopharyngeal carcinoma, IMRT treatment results in improved dose distribution within CTV compared to 3DCRT. Keywords: Nasopharyngeal carcinoma, 3DCRT, IMRT, Integrated boost