An easy method to create the better representative of patients with various body morphologies

Introduction: In recent years, there has been increasing demand for personalized anatomy modeling for medical applications, because the influence of phantoms on the quality of treatments and precision dosimetry has been specified. Phantoms have many applications on medical such as treatment planning, diagnostic imaging, clinical radiological exposure simulation, and biomechanics analysis. The first generation of phantoms was based upon mathematical expressions describing idealized body organs. The second generation was based on three-dimensional images of individuals (CT or MRI images), they offer a more realistic anatomy. This generation was named voxel phantoms. Unfortunately, these phantoms are often limited to a single reference size, which often may not be representative of the patient population at large because they are in the different weight percentile. On the other hand, the construction of specific- patient phantoms is time -consuming and costly. The aim of this study by adding muscle and adipose tissues to a reference adult ORNL phantom torso (50th percentile) was built 65th, 75th, 85th and 95th weight percentiles. To ensure the method, results were acquired for VIPMAN and NORMAN voxel phantoms. The obtained results were compared with that of reported in previous study.
Materials and Methods: The difference between the thickness of the torso muscle and adipose tissues between ORNL with VIPMAN, NORMAN and weight percentiles above the 50th percentile was determined then its equivalent the muscle and adipose tissues were added to the ORNL torso. By appending these layers, the skin and breasts positions were altered. The front and back skin of the torso was also separated and was simulated by two distinct cells in the Monte Carlo code since the tissues added to the front and back of the torso were not the same. Simulations were performed using MCNPX2.4.0 Monte Carlo code for photons with energies 10keV to 10 MeV for anterior-posterior (AP), posterior-anterior (PA), left-lateral (LLAT) and right-lateral (RLAT) irradiation geometries. ENDF/B-VII cross section library was used for calculations. Kerma approximation was applied for energies lower than 500 keV and electron transport was done for energies above 500 keV. Effective dose was calculated using the radiation and tissue weighting factors (wR and wT) from the recommendation of ICRP publication 103.
Results: Differences are reported as the mean relative difference of effective dose and ± SD. The greatest differences for VIPMAN and NORMAN was observed in PA (23.80±12.30) and RLAT (8.20±12.10) respectively. These differences were decreased to 12.10±9.10 and 4.90±2.10 for VIPMAN and NORMAN after adding extra tissues to the ORNL torso. According to results there is a satisfactory agreement between the ORNL, VIPMAN and NORMAN after the addition of the appropriate thickness to the back and front of ORNL torso. Comparisons between different weight percentiles indicate that effective dose was decreased with increasing the weight percentile, so that the mean relative difference of effective dose in 65th, 75th, 85th and 95th with 50th percentile is 15.00±12.04, 18.17±13.05, 20.00±13.93and 24.93±15.17 respectively for AP. As results are shown with increasing percentile, differences and SD were significantly increased.
Conclusion: The results show that by adding extra tissues to an existing 50th percentile can be created other weight percentile. This method can easily be used to expand the phantoms library in order to improve the treatment planning and diagnostic procedure.