System for determining the presence of a substance of interest in a sample
Inventors
Troy, Jerome • Edwards, John W. • Goldman, Heather • Joiner, David • Miller, William
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11769039-B2
Publication Date
2023-09-26
Expiration Date
Abstract
A detection device for detecting the presence of a substance of interest in a sample is described. The device can include a data store comprising executable instructions for at least one convolutional neural network, CNN, configured to process images: and a processor coupled to the data store and configured to execute the instructions to operate the at least one CNN. The detection device can be configured to: obtain spectrometry data, operate a first one of the CNNs to process the spectrometry data to obtain a first CNN output; apply a mask to the spectrometry data to obtain masked data; operate a second one of the CNNs to process the masked data to obtain a second CNN output; and determine if the substance of interest is present in the sample based on both the first CNN output and the second CNN output.
Core Innovation
The disclosed invention relates to a server for analyzing spectrometry data to determine the presence of a substance of interest in a sample. The server includes a data store comprising executable instructions for at least one convolutional neural network configured to process images and a processor coupled to the data store to execute the instructions to operate the at least one CNN. The server is configured to obtain spectrometry data arranged in at least one two dimensional array and process the spectrometry data using CNNs.
After obtaining the spectrometry data, the server operates a first CNN to process the spectrometry data to obtain a first CNN output. The server then applies a mask to the spectrometry data to obtain masked data, and operates a second CNN to process the masked data to obtain a second CNN output. The server determines whether the substance of interest is present in the sample based on both the first CNN output and the second CNN output.
The disclosure further describes generating image-format data for CNN processing from spectrometry array data and supporting multiple spectrometry modality types. It also describes masking mechanisms that reduce intensity or zero intensity of selected regions, and decisioning based on probability and threshold value. The disclosure additionally describes optional voting ensemble behavior using multiple CNNs and masks, and updating CNN instructions based on incorrect determinations.
Claims Coverage
The independent claims are two: a server architecture and a system architecture. Across these independent claims, the core inventive concept uses two CNN passes on spectrometry data—one on unmasked data and one on masked data—followed by determining substance presence based on both CNN outputs.
Server obtains two-dimensional array spectrometry data and analyzes substance presence using first and second CNN outputs
A server configured to obtain spectrometry data arranged in at least one two dimensional array, operate a first CNN to process the spectrometry data to obtain a first CNN output, apply a mask to obtain masked data, operate a second CNN to process the masked data to obtain a second CNN output, and determine if the substance of interest is present based on both the first CNN output and the second CNN output.
System detects substance presence using a spectrometer and a server with first and second CNN outputs
A system comprising a spectrometer and a server, wherein the spectrometer obtains array data representative of the sample and the server obtains spectrometry data for the sample based on the array data; in response to obtaining the spectrometry data, the server operates a first CNN to obtain a first CNN output, applies a mask and operates a second CNN to obtain a second CNN output, and the system determines if the substance of interest is present based on both the first CNN output and the second CNN output.
Both independent claims cover determining presence of a substance of interest by running a first CNN on unmasked spectrometry data and a second CNN on masked spectrometry data, and using both CNN outputs to make the presence determination.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Not explicitly described in patent.
Interested in licensing this patent?