irradiation with open radionuclides

odecrackAI and Robotics

Oct 29, 2013 (3 years and 7 months ago)

104 views

R. Runge
1
, M. Wendisch
1
, G. Wunderlich
1
, D.Roggenbuck
2
, R. Hiemann
3
, U. Kasten
-
Pisula
4
, K. Storch
5
, J. Kotzerke
1


1
Klinik und Poliklinik für Nuklearmedizin, Universitätsklinikum C.G. Carus an der TU Dresden,
2
GA Generic Assays GmbH,
Dahlewitz

3

Bio
-

Chemie
-

und Verfahrenstechnik, Fachhochschule Lausitz,
4
Klinik und Poliklinik für Strahlenbiologie und Radioonkologie, Universitätsklinikum
-
Hamburg
-
Eppendorf

5
OncoRay
, Zentrum für Strahlenforschung in der Onkologie, Medizinische Fakultät, TU Dresden

The

measurement

of

DNA

double

strand

breaks

(DNA
-
DSB)

with

g
H
2
AX
-
immunof luorescence

microscopy

(
g
H
2
AX
-
IFM)

is

an

established

method

of

detecting

DNA

damage
,

f ollowing

cellular

exposure

to

ionizing

radiat ion
.

The

v isual

count ing

of

the

f oci

is

time

consuming

as

well

as

objectiv e

and

subject

to

art ef act s
.

Theref ore

t here

is

interest

in

an

automated

method
,

which

also

allows

for

standardization

of

the

assessment
.

With

intelligent

computer

based

pattern

recognition
,

inter
-

and

intra
-
laboratory

v ariance

can

be

reduced
.




The

slides

were

prepared

f or

laboratory

internal
,

external

and

automated

analy sis

through

irradiation

of

PC

Cl
3

t hy roid

cells

wit h

t he

bet a
-
emitt er

188
Re

(I TG,

Munich
)

in

a

dosage

range

of

0
-
5

Gy

(dose
-
point
-
kernels
)
.

Immidiately

after

irradiation

the

cells

were

f ixed
,

permeabilized

and

incubated

with

primary

and

secondary

antibodies

(
anti
-
phospho
-
histone

H
2
A
.
X,

Alexa

Fluor

488
)
.

The

slildes

were

prepared

f or

5
-
f old

determination
,

and

distributed

to

t hree

laborat ories

as

blind

samples

f or

visual

counting
.

Laboratory

1
:

Univ ersity

Hospital

Dresden,

Nuclear

Medicine
,

3

inv estigators
;

Laboratory

2
:

Univ ersity

Hospital

Hamburg
-
Eppendorf,

1

inv estigator
;

Laboratory

3
:

TU

Dresden,

OncoRay
,

1

investigator
)
.

T
he

automated

measurement

was

perf ormed

with

the

Aklides
®
-
System

(
Medipan
,

Dahlewitz
)
.

Methods

Results


Both

the

v isual

interpretations

of

the

investigators

(
laboratories

1
-
3
)

as

well

as

the

automatic

ev aluation

show

a

dose
-
dependent
,

linear

increase

of

the

foci

count
/
cells

following

irradiation

with
188
Re
.

The

Aklides
®

algorit hms

were

opt imized

to

f ocus

on

counting

and

size

def inition

(
cell

nuclei
,

f oci
)

based

on

digitized

images

f rom

the

three

inv estigators

of

laboratory

1
.

The

results

of

the

three

investigators

of

laboratory

1

showed

low

within
-
laboratory

variation

and

a

good

correlation

with

the

results

of

the

ev aluation

with

Aklides
®
.

Wit h

one

of

the

external

investigators

(
laboratory

3
),

a

signif icantly

higher

number

of

f oci
/
cells

were

counted

with

2

Gy

und

5

Gy

in

comparison

to

laborat ory

1
.

The

result s

of

the

second

external

investigator

(
laboratory

2
)

showed

at

0
-
2

Gy

a

signif icant

inter
-
laboratory

v ariation

and

at

5

Gy

a

high

lev el

of

agreement

with

laboratory

1
.

Conclusions

Figure

1
:

Comparison

of

visual

counts

by

(
A)

three

investigators

from

labor ator y

1

and

(B)

investigators

of

laboratories

1
-
3

with

automated

eval uati on
.

Representation

of

foci
/
cells

following

irradiation

of

PC

Cl
3

cells

with
188
Re
,

means

and

standard

deviations
.

Figure

2
:

Immunofluorescence

microscopic

detecti on

of

doubl e

strand

breaks

in

PCCl
3

cells

f ollowing

irradi ation

with

1
Gy

188
Re

(D
-
F)

as

well

as

i n

uni r r adi at e d

c ont r ol

cel l s
(

A
-
C
)
.

A/D
:

Nucl ear

st ai ni ng

wi t h

DAPI

B/E:
Repr esent at i on

of

g
H2AX
-
Foci
with

AlexaFluor488

C/F:
Overlaying

of

the

fluorescence

images

G
:

Results

of

automated

e v a l u a t i o n


C

T h e

aim

of

this

study

is

the

adaptation

of

the

pattern

recognition

algorithms

of

an

automated

system
,

f or

the

reproducible

analy sis

of

g
H
2
AX
-
Foci
.

In

addit ion

t he

v isual

int erpret at ions

of

f ive

investigators

are

compared

to

t he

aut omat ed

analy sis

of

Aklides
®
.

Visual and
automatic

interpretation

of

g
H2AX
immunfluorescence

microscopy

images

after
irradiation

with

open
radionuclides

The

adaptation

of

the

automatic

evaluation

of

AKLIDES
®

to

t he

v isual

assessment

makes

t he

aut omat ed

ev aluat ion

of

g
H
2
AX
-
Foci

possible
.

When

comparing

the

results

of

diff erent

laboratories

problems

arise
,

due

to

object iv e

(
dif f ering

microscopes
,

t echnique
)

and

subjectiv e

causes

(
expectations

of

the

evaluator
)
.

To

achive

a

reduction

in

inter
-
laboratory

v ariation
,

f urther

standardization

is

necessary
.

A

comparativ e

analysis

based

on

stored

images

is

planned
,

to

minimize

object iv e

and

subjective

error
.

The

Aklides
®

algorithms

were

optimi zed

to

f ocus

on

counting

and

size

def inition

(
cell

nuclei
,

foci
)

on

the

basis

of

digitalized

images

from

three

investigators

from

laboratory

1
.

Specimens

were

taken

systematically

with

a

60
x

objective

(
f luorescencemicroscope

IX
81
,

Oly mpus
)

in

three

z
-
lev els

at

1
µm

intervals
.

The

evaluation

of

the

foci

and

local

intensities

took

place

in

the

z
-
lev els

with

measurements

f rom

the

groups

of

statistical
,

morphometric

and

f orm
-
describing

characteristics
.

Background and Motivation

A

0

5

10

15

20

25

30

35

0

0,5

0,75

1,0

2,0

5,0

Dose [Gy]

Foci/cells

Automated Aklides system

Visual
evaluation

by

laboratory

1

B

0

5

10

15

20

25

30

35

40

0

0,5

0,75

1,0

2,0

5,0

Dose [
Gy
]

Foci
/
cells

Labor atory1

Laboratory 1

Laboratory 1

Aklides

Laboratory 2

Laboratory 3