2016- Current research.

During the first months of my PhD, supervised by Juan José Gómez Cadenas, I am finishing some studies on the CRT (Coincidence Resolving Time) of our new PET system (PETALO) and we are also trying to apply deep learning to our particle physics problems, both in NEXT and PETALO.

For NEXT the first interesting thing to do is discriminating between signal and background, then we could try track reconstruction, localization and energy measurement. We think all these problems can be attack from a machine/deep learning perspective.

In PETALO there are several levels at which machine learning can also be applied. First to distinguish between photoelectric and compton events. In the latter ones, the first vertex could be potentially identified, improving significantly our sensitivity. Second, interaction vertex position can be computed better using (deep?) neural networks than the classic barycenter algorithm. In the end, we also want to apply deep learning at a medical-image level to classify them, identify objects, etc.
I have worked on the data acquisition (DAQ): preprocessing of the signals, decoding the data sent by electronics, testing zero suppression mode, etc.

I also started to study the effect of cosmic muons in our experiment, which could potentially be one contribution to the background. This muons can produce neutrons by different mechanisms which could activate the 136Xe into 137Xe, a beta emitter which could produce some tracks with energy similar to Qββ.

2015. PETALO: Positron Emission Tomography Apparatus based on Liquid xenOn.

For my master thesis (supervised by Juan José Gómez Cadenas) I have analyzed data from simulations of this new PET to study it's energy and spatial resolutions and CRT (Coincidence Resolving Time).

2013-2014. Memory cost of Quantum Contextuality

Quantum contextuality is a property exhibit by quantum systems, implying that the outcome of a measurement depends on which compatible observables has been measured before. This is shown by the Kochen-Specker theorem. In a previous work done by Kleinmann et al. it was studied the amount of memory needed to simulate a two-qubit system (Peres-Mermin square) using finite automata (Mealy machines). If the memory needed violates Holevo's bound it would be a sign that certain hidden variable theories would not be possible. My work consisted in extending this study to a three qubit system (Mermin star), this was done under the supervision of José Ra. Portillo as my undergraduate thesis.

Undergraduate Thesis can be found HERE.