Multi-channel system and methods for sorting particles
A multi-channel system for classifying particles in a mixture of particles according to one or more characteristics including a common source of electromagnetic radiation for producing a beam of electromagnetic radiation and a beam splitter for producing multiple beams of electromagnetic radiation for directing multiple beams of electromagnetic radiation to each interrogation location associated with each flow channel of the multi-channel system.
1. A method of analyzing sperm comprising:
a. detecting fluorescence emissions produced by stained sperm illuminated with a laser beam;
b. producing waveform pulses based on detected fluorescence emissions produced by stained sperm illuminated with the laser beam;
c. extracting features of the waveform pulses in the form of values;
d. representing the values of the extracted features in a feature space;
e. calculating a posteriori probability the stained sperm belongs to at least a first population of X chromosome bearing sperm and a second population of Y chromosome bearing sperm based on positions of the represented extracted feature values in the feature space and a priori probabilities associated with those same positions in the feature space; and
f. classifying individual sperm as belonging to the first population of X chromosome bearing sperm or the second population of Y chromosome bearing sperm based on which population has a higher posteriori probability.
2. The method of claim 1 wherein the feature space comprises values having multiple components with uni-variate Gaussian distributions of extracted features or values having multiple components with bi-variate Gaussian distributions of extracted features.
3. The method of claim 2 wherein the priori probabilities associated with the positions in the feature space are based on a model selected from: a model having two components with Gaussian distributions, a model having three components with Gaussian distributions; and a model having four components with Gaussian distributions.
4. The method of claim 1 , wherein a third population of stained sperm comprises unaligned X- and Y-chromosome bearing sperm.
5. The method of claim 1 , wherein a third population of stained sperm comprises unaligned X-chromosome bearing sperm and a fourth population of stained sperm comprises unaligned Y-chromosome bearing sperm.
6. The method of claim 1 wherein the extracted features include a peak height and a peak area of the waveform pulses produced by stained sperm.
7. The method of claim 1 further comprising a step of establishing a decision boundary at which the posteriori probability of a stained sperm belonging to the first population of X chromosome bearing sperm is equal to the posteriori probability of the stained sperm belonging to the second population of Y chromosome bearing sperm.
8. The method of claim 1 wherein steps e) and f) are only applied to sperm having waveform pulses including extracted features having a pulse width indicative of a single X- or Y-chromosome bearing sperm.
9. The method of claim 1 , wherein the priori probabilities associated with the positions in the feature space are based on a Gaussian mixture model.
10. A method of sorting sperm comprising:
a. detecting fluorescence emissions produced by stained sperm illuminated with a laser beam;
b. producing waveform pules based on detected fluorescence emissions produced by stained sperm illuminated with the laser beam;
c. extracting features of the waveform pulses in the form of values
d. representing the values of the extracted features in a feature space;
e. calculating a posteriori probability the stained sperm belongs to at least a first population of X chromosome bearing sperm and a second population of Y chromosome bearing sperm based on positions of the represented extracted feature values in the feature space and a priori probabilities associated with those same positions in the feature space; and
f. classifying individual sperm as belonging to the first population of X chromosome bearing sperm or the second population of Y chromosome bearing sperm based on which population has a higher posteriori probability; and
g. sorting sperm based on the classification.
11. The method of claim 10 , wherein the feature space comprises values having multiple components with uni-variate Gaussian distributions of extracted features or values having multiple components with bi-variate Gaussian distributions of extracted features.
12. The method of claim 11 , wherein the priori probabilities associated with the positions in the feature space are based on a model selected from: a model having two components with Gaussian distributions, a model having three components with Gaussian distributions; and a model having four components with Gaussian distributions.
13. The method of claim 10 , wherein a third population of stained sperm comprises unaligned X- and Y-chromosome bearing sperm.
14. The method of claim 10 , wherein a third population of stained sperm comprises unaligned X-chromosome bearing sperm and a fourth population of stained sperm comprises unaligned Y-chromosome bearing sperm.
15. The method of claim 10 , wherein the extracted features include a peak height and a peak area of the waveform pulses produced by stained sperm.
16. The method of claim 10 , further comprising a step of establishing a decision boundary at which the posteriori probability of a stained sperm belonging to the first population of X chromosome bearing sperm is equal to the posteriori probability of the stained sperm belonging to the second population of Y chromosome bearing sperm.
17. The method of claim 10 , wherein the step of sorting sperm based on the classification further comprises separating sperm by electromagnetic deflection based on the classification.
18. The method of claim 10 , wherein the step of sorting sperm based on the classification further comprises photo-damaging sperm based on the classification.
19. The method of claim 10 , wherein the priori probabilities associated with the positions in the feature space are based on a Gaussian mixture model.