Methods and systems using mathematical analysis and machine learning to diagnose disease
Exemplified method and system facilitates monitoring and/or evaluation of disease or physiological state using mathematical analysis and machine learning analysis of a biopotential signal collected from a single electrode. The exemplified method and system creates, from data of a singularly measured biopotential signal, via a mathematical operation (i.e., via numeric fractional derivative calculation of the signal in the frequency domain), one or more mathematically-derived biopotential signals (e.g., virtual biopotential signals) that is used in combination with the measured biopotential signals to generate a multi-dimensional phase-space representation of the body (e.g., the heart). By mathematically modulating (e.g., by expanding or contracting) portions of a given biopotential signal, in the frequency domain, the numeric-based operation gives emphasis or de-emphasis to certain measured frequencies of the biopotential signals, which, when coupled with machine learning, facilitates improved diagnostics of certain pathologies.
1. A system comprising:
a storage area network device configured to store a captured biopotential signals of a subject, wherein the captured biopotential signals are transmitted from a biopotential measuring equipment that non-invasively captured, from one or more electrical leads placed on a subject, biopotential signals as a biopotential signal data set, wherein the captured biopotential signals are used to detect one or more diagnosable pathologies of the subject, wherein the detection operation comprises:
generating, via a processor, one or more fractional derivative data sets by numerically performing, for each of a first fractional derivative data set and a second fractional derivative signal data set of the one or more fractional derivative data sets, one or more fractional derivative operations on the biopotential signal data set in a frequency domain and converting a result of the one or more fractional derivative operations to a time domain signal data set, wherein each of the first generated fractional derivative signal data set and the second generated fractional derivative signal data set comprises a same length and a same sampling frequency as the biopotential signal data set; and
generating, via the processor, a three-dimensional space data set wherein each corresponding value of the biopotential signal data set, the first fractional derivative signal data set, and the second fractional derivative signal data set forms a three-dimensional point in said space data set, wherein geometric features and/or dynamical properties of the three-dimensional space data set are used as variables representative of the subject in a machine learning operation to detect one or more diagnosable pathologies of the subject.
2. The system of claim 1 , wherein the first fractional derivative signal data set is generated by performing a first numeric fractional derivative operation of a first pre-defined order value on the biopotential signal data set in the frequency domain and by performing an inversed transformation operation on the first fractional derived signal data set to convert the first fractional derived signal data set to a time domain signal data set.
3. The system of claim 2 , wherein the inversed transformation operation comprises an inversed Fast Fourier Transform operation.
4. The system of claim 1 , wherein the second fractional derivative signal data set is generated by performing a second numeric fractional derivative operation of a second pre-defined order value on the biopotential signal data set and by performing an inversed transformation operation on the second fractional derived signal data to convert the second fractional derived signal data set to a time domain signal data set.
5. The system of claim 1 , wherein each of the first fractional derivative signal data set and the second fractional derivative signal data set comprises a time domain signal data.
6. The system of claim 1 , wherein the first fractional derivative signal data set is generated by performing a fractional derivative operation of an order of pi/2 or an order of 0.5.
7. The system of claim 1 , wherein the geometric features and/or the dynamical properties of the three-dimensional space data set are generated by performing a modified matching pursuit (MMP) algorithm operation on the three-dimensional point in said space data set.
8. The system of claim 1 , wherein the biopotential signal data set is associated with a biopotential signal collected from a single electrical lead.
9. The system of claim 1 , wherein the electrical-lead-collected measurements of the at least one biopotential signal are collected at a location selected from the group consisting of a chest line of the subject, a waistline of the subject, a wrist of the subject, a pelvic line of the subject, a neck of the subject, an ankle of the subject, a forehead of the subject, and an arm line of the subject.
10. The system of claim 1 , wherein the detection operation comprises:
generating, via the processor, an alpha shape or a Delaunay triangle mesh of the three-dimensional point in said space data set, wherein the geometric features and/or the dynamical properties of the three-dimensional space data set include the geometric features of the alpha shape or the Delaunay triangle mesh.
11. The system of claim 1 , wherein the biopotential signal data set comprises electrocardiographic data.
12. The system of claim 1 , wherein the machine learning operation comprises an artificial neural network algorithm or a regression random forest algorithm.
13. The system of claim 1 , wherein the biopotential signal data set are transmitted, over the network, to an analysis system that performs the detection operation.
14. The system of claim 13 , wherein the system comprises a storage area network (SAN).
15. A method comprising:
storing, in a network, at a storage area network (SAN), a biopotential signal data set of biopotential signals captured via a biopotential measuring equipment configured to capture one or more biopotential signals from one or more electrical leads placed on a subject, wherein the stored biopotential signal are analyzed to detect one or more diagnosable pathologies of the subject, wherein the detection operation comprises:
generating, via a processor, a first fractional derivative data set and a second fractional derivative signal data set by numerically performing, for each of the first fractional derivative data set and the second fractional derivative signal data set, one or more fractional derivative operations on the biopotential signal data set in a frequency domain and converting a result of the one or more fractional derivative operations to a time domain signal data set, wherein each of the first generated fractional derivative signal data set and the second generated fractional derivative signal data set comprises a same length and a same sampling frequency as the biopotential signal data set; and
generating, via the processor, a three-dimensional space data set wherein each corresponding value of the biopotential signal data set, the first fractional derivative signal data set, and the second fractional derivative signal data set forms a three-dimensional point in said space data set, wherein geometric features and/or dynamical properties of the three-dimensional space data set are used as variables representative of the subject in a machine learning operation to detect one or more diagnosable pathologies of the subject.
16. The method of claim 15 further comprising:
transmitting, over the network, the stored biopotential signal to an analysis system, wherein the analysis system receives the stored biopotential signal data set and performs the detection operation to detect the one or more diagnosable pathologies of the subject.
17. The method of claim 15 , wherein the one or more diagnosable pathologies include a congenital heart defect (CHD) in a mammal.
18. The method of claim 15 ,
wherein the first fractional derivative signal data set is generated by performing a first numeric fractional derivative operation of a first pre-defined order value on the biopotential signal data set in the frequency domain and by performing an inversed transformation operation on the first fractional derived signal data set to convert the first fractional derived signal data set to a time domain signal data set, and
wherein the second fractional derivative signal data set is generated by performing a second numeric fractional derivative operation of a second pre-defined order value on the biopotential signal data set and by performing an inversed transformation operation on the second fractional derived signal data to convert the second fractional derived signal data set to a time domain signal data set.
19. The method of claim 15 , wherein the electrical-lead-collected measurements of the at least one biopotential signal are collected at a location selected from the group consisting of a chest line of the subject, a waistline of the subject, a wrist of the subject, a pelvic line of the subject, a neck of the subject, an ankle of the subject, a forehead of the subject, and an arm line of the subject.
20. The method of claim 15 ,
wherein the geometric features and/or the dynamical properties of the three-dimensional space data set are generated by performing a modified matching pursuit (MMP) algorithm operation on the three-dimensional point in said space data set, and
wherein the machine learning operation comprises an artificial neural network algorithm or a regression random forest algorithm, and
wherein the detection operation comprises generating, via the processor, an alpha shape or a Delaunay triangle mesh of the three-dimensional point in said space data set, wherein the geometric features and/or the dynamical properties of the three-dimensional space data set include the geometric features of the alpha shape or the Delaunay triangle mesh.