Smart AFMs

Machine learning is already a thing and it is slowly percolating into the most unexpected places. AFM is its most recent victim and Juan F. Gonzalez-Martinez et al. (Biofilms-Research Center for Biointerfaces at Malmö Univerity) have put them together.

Although deep learning techniques had already been used for AFM related analysis, for the first time (to my knowledge) it’s been used to drive the microscope to locate particularities of the samples. They have used Plasmid DNA from E. coli with assubject, and the challenge was to teach the microscope to identify and distinguish single molecules and take images of them with different lateral resolutions.

Although this type of studies is not technically challenging, doing them is extremely time consuming because it needs the close supervision of the researcher. Thus, this breakthrough could open the door not only to a full autonomous AFM but to the analisys of large amounts of samples and statistically significant studies.

We did this picture (featured in the inside cover of Nanoscale) together with Javier Sotres, first author of the paper.


On Wireless BioSensors!

The last paper of T. Ruzgas, J. Sotres etal at Malmö University (Sweden) starts with a disturbing statement: “It is predicted that with the development of Internet of Things technology by 2025 we expect more than 1000 connected devices per human”. With this idea in mind they are studying how to develop robust and cheap biosensors that will provide us with health information. And for that they are exploiting the ability of enzymes to “establish direct electron transfer contact with electrically conducting materials”.


This research, that made it to the cover of ChemElectroChem, is getting us closer to a cyborg-like healthier future.