At Zimmer and Peacock we are a diverse team with skills from biosensors engineering to data science, one of the unique services we bring to our sensors and biosensors is the application of Artificial Intelligence and Machine Learning.
AI can be used to to analyse the raw signal form a biosensor in several ways:
1) Categorize - Does the signal correspond to the analyte of interest. The application of AI can increase the specificity of the sensor. The algorithms can sort the signals into different categories, and ensure that the signal is due to the analyte of interest.
2) Detect Anomalies - When sensors are out in the field then something could go wrong at several levels including, contamination of the same etc. AI can 'look' at the signal and answer the questions 'does the signal look right ?'.
3) Reduce the noise on the sensor - The signal from a sensor and biosensor changes over seconds and minutes, whilst signal interference such as electrical noise can interfere on the sub-second timeline, therefore it is possible to train the AI to distinguish the signal from the noise.
4) Identify patterns - The human brain is very good at spotting patterns in data from a single sensors and also at spotting the patterns in data between sensors. Similarly AI can be trained to do pattern recognition, so that sensor and biosensor data that might otherwise have had to be interpreted by a PhD can be interpreted by an AI system.
When ZP brings an AI strategy into a clients sensor or biosensors program, we discuss and implement a strategy of gathering, organizing and interpreting the data.
We can create a workflow where the data can come from one of three sources or a combination of sources, including:
1) ZP - ZP's engineers and scientists can gather data in our labs or out on location.
2) YOU - Your team can gather the data.
3) YOUR CLIENTS - We can develop an IT workflow where the end users/actual customers are gathering the data and it flows into the databases to improve the AI.