Determining population boundaries using radial density histograms
Apparatuses and methods for determining population boundaries are described. In one embodiment, population boundaries are determined using radial density histograms.
1. A method for identifying a cluster in two-dimensional data in a two-dimensional data space, comprising:
segmenting, via an apparatus including a processor, said data space into a plurality of radial segments originating at an originating point, wherein said plurality is at least three, and said originating point is in said cluster;
for each radial segment, generating a radial density histogram;
determining a cluster boundary from said radial density histogram, to generate a set of cluster boundaries, one for each radial segment; and
identifying said cluster as consisting of a subset of said data that is enclosed by the set of cluster boundaries.
2. A method for identifying a cluster in two-dimensional data in a two-dimensional data space, comprising:
segmenting, via an apparatus including a processor, said data space into a plurality of radial segments originating at an originating point, wherein said plurality is at least three, and said originating point is in said cluster; and
for each radial segment, generating a radial density histogram;
determining a cluster boundary from said radial density histogram, to generate a set of cluster boundaries, one for each radial segment;
constructing a polygonal gate surrounding said cluster from said set of cluster boundaries; and
identifying said cluster as consisting of a subset of said data that is enclosed by said polygonal gate.
3. The method of claim 2 , wherein said polygonal gate is a convex polygonal gate.
4. The method of claim 2 , wherein said data are obtained using a flow cytometer.
5. A non-transitory computer readable medium comprising instructions executable by a processor of an apparatus, the instructions causing the apparatus to identify a cluster in two-dimensional data in a two-dimensional data space by:
segmenting said data space into a plurality of radial segments originating at an originating point, wherein said plurality is at least three, and said originating point is in said cluster;
for each radial segment, generating a radial density histogram;
determining a cluster boundary from said radial density histogram, to generate a set of cluster boundaries, one for each radial segment; and
identifying said cluster as consisting of a subset of said data that is enclosed by the set of cluster boundaries.