Your Responsive Page

Andres M. Arias Lorza

PhD

“Being successful in research requires a combination of knowledge,

creativity, discipline, and social skills.”

About me:

I am from Cali, Colombia (The salsa capital of the world!). I graduated in the year 2000 from Lacordaire high school (top 100 school in Colombia back in the days), where I was one of the best in my class (if not the best) in Mathematics and Physics. Because of this, and that I was always interested as I kid in how things work, I decided to study Electronics Engineering in Javeriana University (top 10 universities in Colombia). During my studies, I became very good in digital electronics, signal processing, and programming.

I did my Bachelors thesis in voice recognition using machine learning (when it was not so cool almost 20 years ago). In my thesis, together with my thesis partner we used voice commands that were recognized by a digital signal processor (a small board with a micro-processor chip among other specialized devices) to control a robotic arm. The thesis was so successful (I got 9/10 score), that after graduation in 2007, I transformed it into a toy product where I used the voice commands to control a radio controlled car. I even got funding support from the BID (Inter-american development bank) to transform the product into a business. During the next two years, I continued working in my business ideas and developed software to manage small businesses (generating and printing the bills, and control the data base of the company among other things), and also searched for masters programs to continue my studies. As I wanted to study in a top world university, I spent some time researching good master programs, doing the applications, and improving my English skills. I even went for three months to do an English course in USA in 2007.

In 2009 I got accepted for the Embedded Systems program in Eindhoven University of Technology (top 100 engineering school in the world). This program was mainly intended to teach advanced topics in hardware and software design, which I was very interested back then. I got an scholarship from Colombia to pay my studies and in the second semester of 2009 I went to study in Eindhoven, The Netherlands. The two years I spent there were great academically, culturally, and socially. I met many inspiring people that made me think things in a completely different perspective. Here I also became very interested in how to apply mathematics and engineering in Medicine. After talking with several professors, I decided to do my masters thesis in an exciting and relatively new field called medical imaging. In my thesis I developed algorithms to analyse blood flow using MRI images supervised by the brilliant mathematician Prof. Remco Duits.

I continued my research career for a PhD in medical imaging at Erasmus University (top 100 medical schools in the world) in 2011. For the next four years I developed algorithms for segmentation and registration of arteries using medical images. During this time I was tutored by the legends in field: Prof. Marleen de Bruijne and Prof. Wiro Niessen. The PhD made me an expert in medical imaging, I published several papers in top journals, went to speak at several conferences, and got often invited as reviewer. (See here my PhD defence: Link).

In 2018, I started a post-doc in cancer imaging. I really wanted to work with the cancer disease, and this was a great opportunity to work in a fantastic cancer hospital: Moffit Cancer Center (Top 10 cancer hospitals in USA), and to work with a very knowledgeable expert in cancer MRI imaging such as Dr. Raghunand Natarajan. Our team was part of the group of one of the world leaders in cancer imaging: Prof. Robert Gillies (RIP), who was called the father of radiomics. Here I worked in predictors of novel cancer treatments. Unfortunately, COVID came, I worked from home for about 2 years and my working visa expired. This is what Dr. Raghunand Natarajan said about me: Link.

From May 2022 until October 2023, I worked as associate researcher at Zuse Institute Berlin, an institution for mathematics and informatics research located inside the world renown Free University of Berlin. Here I helped in training some of the PhD students and did research related to deep learning for image segmentation and image reconstruction.

My interests are to continue working in solving problems where I can use both my strong mathematics and technical skills. Medical problems such as cancer research that would require the use of data analysis, imaging, and machine learning are of special interest to me, therefore feel free to contact me if you want to fund my research. Also contact me if you need consultancy in my topics of expertise related to machine learning, computer vision, and medical imaging. I would also be happy to train a group of engineers or researchers, and teaching students in courses such as mathematics, programming, computer vision, medical imaging, and machine learning. Also contact me to review a report or research paper in a related topic.

Download my CV . Descargar Hoja de Vida en español.

Projects:

Imaging predictors of cancer treatments: In this project I extracted quantitative parameter maps from the MRI images using the method I developed illustrated in the figure below (the developed Matlab tool can be downloaded from my Github site). Parameter maps for a liver tumour for different visits are shown in the next figure. From the figure, changes can be appreciated before (baseline) and after the drug. This information can be used to analyse drug function and to predict if a patient is going to respond to the drug. This project was developed together with Dr. Natarajan Raghunand at Moffitt Cancer Center. Also multiple publications at top journals were written.
- A. Arias Lorza et al. “Tumor Response to Stroma-Modifying Therapy: Magnetic Resonance Imaging Findings in Early-Phase Clinical Trials of Pegvorhyaluronidase alpha (PEGPH20)”, 2023
- A. Arias Lorza et al. “Dose-response assessment by quantitative MRI in a phase 1 clinical study of the anti-cancer vascular disrupting agent crolibulin”, 2020
High-resolution reconstruction of dynamic knee MRI: MRI is either acquired at static positions at higher quality and high resolution during a long acquisition or dynamically (several MRI acquired during movement) with multiple images acquired fast at lower quality and low resolution. To improve the quality of the dynamic images, reconstruction techniques are used. At ZIB Berlin, I developed an image reconstruction method for the knee movement based on image registration described in the figure below. Image registration is the process to align the structures of two or more images, in this method I align the knee structures from few static images to the dynamic images considering the elasticity of soft tissues and rigidity of bones. Dynamic images and the reconstruction are shown in the next figure.
- A. Arias Lorza et al. “Image Registration of Dynamic MRI for Strain Analysis of Soft Tissue Structures in the Knee Joint using Elastix”, 2023
Image analysis of arteries: Segmentation of the artery wall is necessary for diagnosis and progression of several diseases such as atherosclerosis. Multi-modality imaging of the artery is also necessary to analyse several properties of the artery wall. At Erasmus Medical Center, together with Prof. Marleen de Bruijne and Prof. Wiro Niessen, I developed a segmentation method of the artery wall in MRI and Ultrasound, and a method to register different image modalities. In the figure below it is described a segmentation of the artery wall overlaying registered MRI and Ultrasound images. This project resulted in many publications in methodological and medical journals.

About me:

Projects:

Contact: