2 - motivate fluorescence microscopy
3 - explain its drawbacks (usually only 5s exposure without changing biology)
6 # /data/floh-junk/0117/final/kielhorn-ssi2011_0120.pdf
7 - distance between consecutive lenses always sum of both focal lengths
8 - source is output of integrator rod
9 - if mirrors are flat, image of source on B1
10 - if mirrors are tilted, light no longer passes through the aperture
11 - mma is imaged into bfp, tilted mirrors are dark there
12 #+CAPTION: Thin lens model of the optical system. L{1..5} lenses, M{1,2,3} mirrors, B1 adjustable aperture, PBS polarizing beam splitter, D1 dichroic beam splitter
14 #+CAPTION: Photograph and drawing of the system.
16 - give examples how it can be used to improve illumination
17 for c. elegans samples
19 - first order in b1 is at an angle FIXME
20 - aperture stop should have FIXME diameter
21 - ideally half orders should be blocked
22 - this leads to an image diameter on the lcos, in the sample plane FIXME
23 ** light homogenization
24 - fiber illuminated under an angle
25 - rotating microlenses with 5 degree divergence FIXME
26 - 250mm light tunnel, needs to be rectangular for mixing effect,
27 caustics in round tunnels prevent homogenization FIXME paper
28 ** Other approaches of light control
29 - other approaches in the literature:
32 - CLEM, feedback loop paper
34 - comparison of the different approaches
35 - SPIM and HiLo only do angular control
36 - SPIM doesn't seem to have any real disadvantage, its sectioning (1 .. 3 um?)
37 isn't as good as CLEM but usually sufficient for embryos
38 - CLEM only works with coherent light and can do good sectioning with a confocal
39 - Levoy spatial resolution is limited and fixed (depending on what
40 microlenses he uses), our system has high spatial resolution but
41 we shouldn't make use of it otherwise the angular control
43 - Levoy is the only one, that can instantaneously illuminate
44 different regions of the sample with different angles. We (and
45 Holo and Kino) could multiplex in time, though.
46 - Holo is an approach that we tried using a different phase
47 gratings in the image plane.
48 - Kino would be an approach where several phase SLM would act as a
49 Kinoform element and redistribute light from dark areas into the
50 bright ones (hasn't been tried to my knowledge).
53 | Approach | Angular | Spatial | Sectioning | Throughput | Incoherent | Advantage | Disadvantage |
54 |----------+---------+---------+------------+------------+------------+------------------+-----------------|
55 | SPIM | ++ | - | + | ++ | + | even drosophilia | sample mount |
56 | HiLo | ++ | - | o | o ? | + | | untilt |
57 | CLEM | - | ++ | ++ | ++ | - | | |
58 | Levoy | + | + | - .. o ? | o | ++ | | adaptive |
59 | Holo | + | + | - | - | -- | only one SLM | discrete angles |
60 | Kino | + | + | - | + or ++ ? | -- | | |
61 | Our | + | ++ | - .. ++ ? | - | - | | adaptive |
63 ** The electronic system
64 - explain how devices are synchronized
65 *** Liquid crystal on Silicon Display
66 - maybe describe how it works (but there isn't much information available)
67 *** Micro mirror array
68 #+CAPTION: MMA images, left: SEM image, middle: optical reflective , right: exaggerated rendering of how a 8x8 checker pattern would be displayed on the device
70 - 256x256 mirrors with 16 um pitch
71 - angle continously adjustable from 0 to 2 degree (blazed condition
72 for full visible range)
73 - use fourier filter to convert into intensity device
74 - Fraunhofer contrast measurement
75 - tilt all mirrors, keep frame flat
76 - measure intensity at three areas in the filtered image
77 - plot intensity versus deflection, find minimum
79 ** The optimization system
80 *** Illumination optimization via Raytracing
81 *** Expected improvement with controlled light exposure (lack of Simulation will make that hard to write)
82 *** Expected improvement with angular illumination
83 *** Experimental results
86 ** Simulating the microscope
87 *** Raytracing through immersion objective
88 *** On locating nuclei
89 *** On tracking nuclei
90 *** Angular point spread function
91 *** Comparison of widefield and spatio-angular illumination
95 function inten=memi_inten(px,py,mpx,mpy,mr,lpx,lpy,lr,h_asf,filtersize)
96 Ex=squeeze(h_asf(:,:,:,0));
97 Ey=squeeze(h_asf(:,:,:,1));
98 Ez=squeeze(h_asf(:,:,:,2));
101 sx=size(h_asf,1);sy=size(h_asf,2);sz=size(h_asf,3);
109 t=sqrt((xx(sx,sy)-mpx).^2+(yy(sx,sy)-mpy).^2)<mr;
111 t=gaussf(t,filtersize);
113 lcos_ill=ift(ft_tun.*t);
115 L=sqrt((xx(sx,sy)-lpx).^2+(yy(sx,sy)-lpy).^2)<lr;
116 lcos_out=lcos_ill.*L;
117 %% Apply ASF to Amplitude after LCOS
118 lcos_3D=newim(sx,sy,sz,'complex');
119 lcos_3D(:,:,floor(sz/2))=lcos_out;
120 Esx=convolve(lcos_3D,Ex);
121 Esy=convolve(lcos_3D,Ey);
122 Esz=convolve(lcos_3D,Ez);
123 inten=real(Esx*conj(Esx)+Esy*conj(Esy)+Esz*conj(Esz));
133 h_asf=kSimPSF( {'na',1.38;'ri',1.52;'sX',sx;'sY',sy;'sZ',sz;'scaleX',80;'scaleY',80;'scaleZ',160;'computeASF',1;'circPol',0;'scalarTheory',0})
134 h_asf=squeeze(h_asf);
140 inten=memi_inten(px,py,mpx,mpy,mr,lpx,lpy,lr,h_asf,1);
142 fprintf('px: %d, py: %d\n',px,py);
147 *** Optimization of illumination
148 ** Camera calibration
149 compare performance of different cameras (SNR agains light
150 intensity), support our choice of camera
151 ** On measuring light intensity in the focal plane
152 ** On the choice of the objective
153 explain our choice of the objective
154 ** Controlling the micro-mirror array
155 ** Synchronizing electronics by microcontroller
156 ** Preparing C. elegans embryos for imaging