Back

Automatic crystal identification for crystallography: a comparison between direct methods and artificial intelligence strategies

Lages, C. F.; Nicoleti, V. Y.; Correa, M. R.; Mausbach, P.; Ramos, F. C.; Araujo, E.; Nascimento, A. Z.; Miqueles, E. X.

2024-08-23 bioinformatics
10.1101/2024.08.22.609235 bioRxiv
Show abstract

Crystallography is an important, and well-established technique for assessing crystalline atomic structure of a wide variety of materials. An auxiliary optical microscope instrument is used to perform the data acquisition at the manaca beamline, a crystallography experimental station from sirius (the 4th generation Brazilian synchrotron). The importance of detecting these crystals in the auxiliary image is to define the best order in which each crystal will be measured by the incident beam since each one of them must be measured only once; improving beamtime and maximizing user experience at the facility. Detecting rough crystal positions from these auxiliary images is a scientific computing task, which could benefit from some open artificial intelligence tools. This manuscript compares a real-time object detection interface (Yolo) with a new, simple and effective strategy obtained from a particular partial differential equation.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.