A major field of study in science, that have made the understanding of the world and made homo sapiens
the dominant species on the planet is our understanding of medicine. Under this vast umbrella is the
detection and classification of microorganisms. Electrochemical biosensors have emerged as
promising tools in this regard due to their high sensitivity, selectivity, and potential for miniaturization.
By integrating a Machine Learning & Deep Learning model for the detection of the bacterium in the
sample and also to see how different physical and chemical environmental characteristics affect the
working of the sensor. The working of the project is quite elementary, as a sensor is simulated on
COMSOL Multiphysics, in a particular environment, with varying factors of Sensor factors, namely:
Electrode Shape, Material & height, material of the substrate and the dielectric used. This is done to test
varying factors and to find the best combination of materials that produce the most consistent and
accurate values. Following which, the data generated from the simulation, which will be a set of electric
field values can be processed and mutated to the required format and will be fed into the machine
learning algorithm to be trained. The algorithm best suited is the Random Forest Method, where a
decision tree is formed to detect the bacteria present in the sample. Further, a Deep Learning model
can also be trained to add a layer of complexity, which shows how each of the underlying chemical &
physical properties of different bacterium play a role in generating the final peak electric field values.
This works as the chemical & physical properties is unique to every bacterium. If implemented and
developed, this project can act as a lifesaver in disease detection as “Prevention is better than cure”.
This can correctly can help communities, especially low- income and underprivileged ones. The
applications include, but are not limited to Medical Diagnostics, Environmental Monitoring, Food
Safety, Agriculture, Biotechnology & Pharmaceuticals and Security and Defense.

🔒 This is a subscription article

Full text is available to subscribers and institutional members. Please choose an option below to access it.

A major field of study in science, that have made the understanding of the world and made homo sapiens
the dominant species on the planet is our understanding of medicine. Under this vast umbrella is the
detection and classification of microorganisms. Electrochemical biosensors have emerged as
promising tools in this regard due to their high sensitivity, selectivity, and potential for miniaturization.
By integrating a Machine Learning & Deep Learning model for the detection of the bacterium in the
sample and also to see how different physical and chemical environmental characteristics affect the
working of the sensor. The working of the project is quite elementary, as a sensor is simulated on
COMSOL Multiphysics, in a particular environment, with varying factors of Sensor factors, namely:
Electrode Shape, Material & height, material of the substrate and the dielectric used. This is done to test
varying factors and to find the best combination of materials that produce the most consistent and
accurate values. Following which, the data generated from the simulation, which will be a set of electric
field values can be processed and mutated to the required format and will be fed into the machine
learning algorithm to be trained. The algorithm best suited is the Random Forest Method, where a
decision tree is formed to detect the bacteria present in the sample. Further, a Deep Learning model
can also be trained to add a layer of complexity, which shows how each of the underlying chemical &
physical properties of different bacterium play a role in generating the final peak electric field values.
This works as the chemical & physical properties is unique to every bacterium. If implemented and
developed, this project can act as a lifesaver in disease detection as “Prevention is better than cure”.
This can correctly can help communities, especially low- income and underprivileged ones. The
applications include, but are not limited to Medical Diagnostics, Environmental Monitoring, Food
Safety, Agriculture, Biotechnology & Pharmaceuticals and Security and Defense.

🔒 This is a subscription article

Full text is available to subscribers and institutional members. Please choose an option below to access it.