图1.三种高密度分子定位方法的数据处理结果比较
Figure1.EvaluationofthelocalizationperformanceamongQC-STORM,ThunderSTORMandWindSTORMusingexperimental2Dimages.
Newadvanceinsuper-resolutionfluorescencemicroscopy-Real-timemaximumlikelihoodfittingofmultipleemittersusingthedivideandconquerstrategy
byBrittonChanceCenterforBiomedicalPhotonics
Super-resolutionfluorescencemicroscopyisoneofthemostimportantbreakthroughsinopticalmicroscopyinthe21stcentury,andisawardedwiththeNobelPrizeforChemistryin2014.Super-resolutionlocalizationmicroscopy(SRLM)isarepresentativesuper-resolutionimagingtechnique.Withthecombinationofsingle-moleculefluorescencemicroscopyandhigh-precisionmoleculelocalization,SRLMisabletouseasimpleopticalset-uptoprovide20~30nmspatialresolutionthatenablesunprecedentedopportunitiesforbiomedicalresearches.Inrecentyears,researchershavewidelyagreedthatmulti-emitterlocalizationhasgreatpotentialformaximizingtheimagingspeedofsuper-resolutionlocalizationmicroscopy.However,theslowimageanalysisspeedofreportedmulti-emitterlocalizationalgorithmslimitstheirusageinmostlyoff-lineimageprocessingwithsmallimagesize.
UndertheguidanceofProf.Zhen-LiHuangfromBrittonChanceCenterforBiomedicalPhotonics,WuhanNationalLaboratoryforOptoelectronics,LuchangLi(aPhDcandidateinHuang’sgroup)andothergroupmembersadoptedthewell-knowndivideandconquerstrategyincomputerscienceandpresentedafitting-basedmethodcalledQC-STORMforfastmulti-emitterlocalization.QC-STORMachieves3-4ordersofmagnitudeintheimageprocessingspeedthanthepopularfitting-basedThunderSTORMandtheup-to-datenon-iterativeWindSTORM,withsimilarlocalizationprecisionanddetectionrate.Therefore,QC-STORMiscapableofprovidingreal-timefullimageprocessingonrawimageswith100μm×100μmfieldofviewand10msexposuretime.ThisstudyisreportedrecentlyinOpticsExpress,22ndJuly,2019,Vol.27,Iss.15,pp.21029-21049.