Research

Cancer is the second most important cause of death among non-communicable diseases, and its incidence is expected to rise in the coming decades. Of a hundred deaths that occurred in Chile in 1960, eight were due to malignant tumors. Today, that number rise up to 25 out 100. Moreover, it is estimated by 2020, one in three will die for this reason.

Considering the total number of deaths from all causes, three times more people die from cancer than 50 years ago, which undoubtedly evidences the need for researchers to contribute in this area. Therefore I decided to search for a career involving cancer research.

I turned my programming expertise toward biomedical applications in the areas of radiotherapy, my research has focused primarily on Monte Carlo simulations, the creation of an open source radiation treatment planning system with different modalities (photon, proton and carbon ion) and the time efficiency of dose calculation as well.

Interests

  • Cancer research
  • Radiation therapy
  • Radiation treatment planning system
  • Dose calculation
  • Monte Carlo techniques
  • IMRT, IMPT
  • Ion therapy
  • Parallel computing

Research Projects

  • matRad, open source treatment planning system (TPS)

    Open source TPS for radiation treatment planning of intensity-modulated photon, proton, and carbon ion therapy for educational and research purposes.


     

    matRad is an open source software for three-dimensional radiation treatment planning of intensity-modulated photon, proton, and carbon ion therapy. matRad is developed for educational and research purposes; it is entirely written in MATLAB.  The toolkit features a highly modular design with a set of individual functions modeling the entire treatment planning workflow based on a segmented patient CT. All algorithms, e.g. for ray tracing, photon/proton/carbon dose calculation, fluence optimization, and multileaf collimator sequencing, follow well-established approaches and operate on clinically adequate voxel and bixel resolution. Patient data, as well as base data for all required computations, is included in matRad.

    I developed matRad as a part of my master thesis at the research group Optimization Algorithms within the Division of Medical Physics in Radiation Oncology at the German Cancer Research Center DKFZ in Heidelberg under the supervision of Ph.D. Mark Bangert. These days, matRad’s development is constantly improving it, adding new features and techniques in radiotherapy planning.

    There was an article published by El Mercurio newspaper about matRad

    For further information, please visit http://www.matrad.org or matRad’s publications.

  • CBCT X-ray Monte Carlo simulation

    Characterize the X-Ray Cone Beam Computed Tomography unit (CBCT), which is mounted on the Elekta Synergy Linear Accelerator using Monte Carlo dose calculations techniques.

    The Image Guided Radiotherapy IGRT technologies are becoming more progressively used in a clinical environment, this would lead one important issue: The increased use of x-rays could add significant radiation dose to patients. This would increase the risk of secondary cancer. Therefore there is an increasing interest in estimate the additional cumulative dose received by patients undergoing diagnostic x-ray exposures.

    The Monte Carlo MC method is the ideal tool, which is the most accurate method to model the transport of radiation, therefore is possible to obtain an accurate estimation of the dose given to patients scanned with diagnostic imaging devices like CBCT, especially when in vivo dose measurements are difficult to take. To perform the simulation, I used the EGSnrc Monte Carlo software

    I carried out this simulation as a part of my undergraduate thesis under the supervision of prof. Ph.D. Edgardo Dörner and prof. Ph.D Beatriz Sánchez at the Faculty of Physics – PUC.

  • Monte Carlo parallel computing

    EGSnrc and BEAMnrc parallel computing

    The Monte Carlo (MC) method is considered the most exact approach to calculating dose distributions in radiotherapy. However, due to the long calculation times usually involved this method is considered too slow for routine clinical applications. One usual approach is to use parallelization techniques under various processing units in order to provide reasonable calculation times.

    One of the most widely used MC codes in the field of medical physics is the EGSnrc MP multiplatform environment. In order to introduce parallelism, the EGSnrc MP platform divides the MC calculation into several jobs and submits them to a certain number of machines using a Batch-Queueing-System (BQS).

    This research made possible to Faculty of Physics at Pontificia Universidad Católica de Chile be able to buy two supercomputers with 128 cores each one to perform Monte Carlo simulations, Reducing computation times, benefiting both researchers and students.