figures/

NanoJ: High-Performance Open-Source

Super-Resolution Microscopy in ImageJ
R.F. Laine1-2 *, K.L. Tosheva1 *, R.D.M. Gray1, P. Almada1, D. Albrecht1, J. Mercer1, C. Leterrier3,
P.M. Pereira1-2 , S. Culley1-2 and R. Henriques1-2 ; *equal contribution
1University College London, UK; 2The Francis Crick Institute, UK; 3Aix-Marseille Univ, CNRS, INP, France
Correspondence to: [email protected], [email protected], [email protected]
Poster DOI: 10.6084/m9.figshare.7963556 [width=10cm]c1
‡ github.com/HenriquesLab — 7 @HenriquesLab [width=16cm]c1

Our team has built an open-source image analysis framework for super- Drift Multi-
resolution microscopy designed to combine high performance and ease cor- colour
of use. We named it NanoJ - a reference to the popular ImageJ soft- rec- chan-
tion nel
ware it was developed for. Here we highlight the current capabilities with reg-
of NanoJ for several essential processing steps: spatio-temporal align- NanoJ- is-
ment of raw data (NanoJ-Core), super-resolution image reconstruction Core. tra-
(NanoJ-SRRF), image quality assessment (NanoJ-SQUIRREL), struc- a) tion
tural modelling (NanoJ-VirusMapper) and control of the sample environ- Com- with
pos- NanoJ-
ment (NanoJ-Fluidics). We expect to expand NanoJ in the future through ite Core.
the development of new tools designed to improve quantitative data anal- im- a)
ysis and measure the reliability of fluorescent microscopy studies. age Com-
of pos-
two ite
frames im-
from age
a of
time- multi-
lapse colour
dataset beads,
of prior
the to
same (left)
field- and
of- af-
view. ter
b) (right)
Cross- chan-
correlation nel
map reg-
[width=0.8]c1 (CCM) is-
be- tra-
tween tion
the us-
two ing
frames NanoJ-
shown Core.
in Scale
a). bars:
The 25
vec- m,
tor in-
po- sets:
si- 0.5
tion m.
of b)
the Vec-
max- to-
i- rial
mum rep-
in- re-
di- sen-
cates ta-
the tion
lin- of
ear the
shift shift
be- be-
tween tween
the the
two two
frames. chan-
c) nels
Over- (left,
lay dis-
of place-
the ment
two vec-
frames tor
af- length
ter 50
drift times
cor- larger
rec- for
NanoJ-Core: Drift Correction tion
us-
rep-
re-
ing sen-
NanoJ- ta-
Core. tion
d) pur-
Ver- poses),
ti- hor-
cal i-
and zon-
hor- tal
i- (mid-
zon- dle)
tal and
Drift estimation is achieved by directly analysing a raw sequence of time-
drift ver-
points. Frames can then be directly translated to correct drift or a table curves ti-
with drift coordinates can be generated and used in other algorithms (e.g. ob- cal
SRRF). tained (right)
us- shift
ing maps
NanoJ- ob-
Core tained
from and
the ap-
100- plied
frame to
raw the
data. data
shown
in
NanoJ-Core: Channel Registration a).
Scale
bars:
Channel registration is able to calculate and apply an elastic transform to 25
realign different captured wavelengths. m.

[width=0.8]c1 [width=0.8]c1
NanoJ-SRRF: Live-Cell Super-Resolution Qual-
ity
NanoJ-VirusMapper: Structural Mapping and
and Modelling
res-
o-
lu-
tion
Super-Resolution Radial Fluctuations (SRRF) is able to generate a super- as-
resolution reconstruction by analysing fluctuations in the emission of flu- sess-
ment
orophores captured in a short image sequence. with
NanoJ-
VirusMapper features a single-particle analysis (SPA) algorithm, capable
SQUIRREL.
a) of generating structural models of conserved structures imaged by Super-
A Resolution Microscopy
super-
resolution
ren-
der-
[width=0.85]c1 ing
and
Live- ac-
cell quired
super- wide-
resolution field
mi- im-
croscopy age
with of
fixed [width=0.7]c1
NanoJ-
SRRF. Alexa647
Quan-
a) la-
ti-
Com- belled
ta-
par- mi-
tive
i- cro-
SPA-
son tubules.
based
of b)
mod-
wide- Left:
elling.
field SQUIR-
Mul-
and REL
ti-
SRRF er-
com-
re- ror
po-
con- map
nent
struc- high-
model
tion light-
of
from ing
the
UtrCH- dis-
Vac-
GFP crep-
cinia
actin an-
virus
la- cies
by
belling. be-
imag-
Scale tween
ing
bar: the
in
5 super-
super-
m. resolution
resolution
b) and
hun-
Time- diffraction-
dreds
course limited
of
of im-
flu-
the ages
o-
in- in
res-
set (a).
cently
shown Right:
la-
in Mag-
belled
a), ni-
viruses
ob- fied
and
tained in-
mod-
at sets
elling
33.3 at
their
Hz in-
struc-
and di-
ture
dis- cated
through
played po-
VirusMap-
ev- si-
per
ery tions
30 on
s. er-
Scale ror
bar: map.
1 c)
m. Left:
c) SQUIR-
Colour- REL
coded res-
time o-
course. lu-
Scale tion
bar: map
1 of
m. the
super-
resolution
im-
age
in
(a).
Right:
Mag-
ni-
fied
in-
sets
for
in-
di- NanoJ-Fluidics: Liquid Exchange
cated
res-
o-
lu-
tion
NanoJ-SQUIRREL: Estimating Image Quality blocks.
Whole
& Resolution im-
age NanoJ-Fluidics (Pumpy McPumpface) entails the control of a simple in-
scale
bars expensive LEGO-based syringe pump array that automates live-to-fixed
= imaging and sequential labelling of the sample. Protocols can be run au-
5 tomatically in parallel to the acquisition.
m,
SQUIRREL quantitatively calculates local quality and resolution in in-
super-resolution images. It highlights limitations on the data collected set
and a reference point to help researchers improve imaging fidelity. scale
bars
=
1
m.

[width=0.85]c1 [width=1]c1
Schemat-
ics
Author representation
of
the
NanoJ-
Fluidics
sys-
tem.
a)
3D
side
view
of
a
sin- I recognise a gender inequality in the author list (2F/8M). I am
gle [width=0.8]c1
committed to achieve a better balanced representation of gender
sy- and minorities in our future work.
ringe
pump.
b)
2D
top
view
of
a
sy-
ringe
pump
ar-
ray
(rep-
re-
sent-
ing
4
pumps
out
of
128
max-
i-
mum)
and
a
fluid
ex-
trac-
tion
peri-
staltic
pump,
both
con-
trolled
by
an
Ar- Funded by
duino
UNO.
c)
Ex-
am-
ple
of
pos-
si-
ble
work-
flows [width=0.95]c1