UOG Journal Club: Intra- and interoperator reliability of manual and semi-automated measurements of intracranial translucency

UOG Journal Club: February 2012Intra- and interoperator reliability of manual and semi-automated measurements of intracranial translucency

K. Karl, K. O. Kagan and R. ChaouiVolume 39, Issue 2, Date: February 2012, pages 164–168

Journal Club slides prepared by Dr Asma Khalil(UOG Editor for Trainees)

Spina BifidaSpina Bifida

Rankin J et al., Paediatr Perinat Epidemiol 2000Centre for Disease Control and Prevention statistics 2011

Boyd PA et al., J Med Screen 2011

3-6 per 10,000 births

0

5

10

15

20

25

1930s

Bir

th p

reva

len

ce (

/10,

00

0)

0

5

10

15

20

25

Year

Bir

th p

reva

len

ce (

/10

,000

)

1930s1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

• Prenatal diagnosisPrenatal diagnosis• Folic acid fortificationFolic acid fortification

3–6 per 10,000 births

History5%

MS AFP75%

Ultrasound* 98%

Prenatal DetectionPrenatal Detection

Detection rate by ultrasound at 11-14 weeks is 14%Detection rate by ultrasound at 11-14 weeks is 14%

Open spina BifidaOpen spina Bifida

Wald N et al., Lancet 1974Nicolaides KH et al., Lancet 1986

Campbell J et al., Obstet Gynecol 1987Syngelaki A et al., Prenat Diagn 2011

*2nd trimester

MS AFP, maternal serum alpha fetoprotein.

First-trimester detection of open spina bifida using intracranial translucency (IT)

NormalNormal Open spina bifidaOpen spina bifida

BS,brainstem; BSOB, brainstem–occipital bone distance; f.CM, future cisterna magna; OB,occipital bone; Th, Thalamus.

Chaoui R et al., UOG 2011

Chaoui R et al., UOG 2009

B, brainstem;T, thalamus;M, midbrain; My, medulla oblongata.

Chaoui R et al., UOG 2009

NormalNormal Open spina bifidaOpen spina bifida

Intra- and interoperator reliability of manual and semi-automated measurements of intracranial translucency

Karl et al., UOG 2012

Objective

Examine whether the assessment of IT at 11–13 weeks can be further standardized by using the semi-automated NT algorithm, SonoNT®, in comparison to manual measurement.

Retrospective; 116 stored images at 11–13 weeks; IT measurement

Methodology

• 11–13 weeks• Mid-sagittal plane• Dorsoposterior position • Two experienced operators • Twice manually and twice

using the semi-automated software (SonoNT®)

• Intraoperator and interoperator repeatability

BS,brainstem; f.CM, future cisterna magna; OB,occipital bone; Plex.chor., choroid plexus; Th, Thalamus.

The mean of the two manual measurements of the more experienced operator 2 was considered the ‘gold standard’.

ManualManual Semi-automatedSemi-automated

• Magnify the image

• Place the cursor on the inner border of the echogenic posterior brainstem border and on the inner border of the 4th ventricle choroid plexus

•The longest vertical distance in the middle portion

Methodology - manual measurement Methodology - manual measurement

• Corresponding ‘edge enhancement image’ that reflects the Corresponding ‘edge enhancement image’ that reflects the differencesdifferences in brightness rather than the brightness itself in brightness rather than the brightness itself • Define the echogenic lines delineating the translucencyDefine the echogenic lines delineating the translucency• Calipers placed automatically on the inner borders (inner–inner)Calipers placed automatically on the inner borders (inner–inner)

Methodology – semi-automatedMethodology – semi-automated

Left and central images show semi-automated measurement of nuchal translucency, Moratalla J et al., UOG 2010

•The measurement algorithm connects every point on one ofThe measurement algorithm connects every point on one of the two echogenic lines to all points on the other line, then the two echogenic lines to all points on the other line, then selects the minimum distance for each point selects the minimum distance for each point• From these minimum distances, it selects the largest asFrom these minimum distances, it selects the largest as the final translucency measurement the final translucency measurement

Methodology – semi-automatedMethodology – semi-automated

Left and central images show semi-automated measurement of nuchal translucency, Moratalla J et al., UOG 2010

• IT has a more complex shapeIT has a more complex shape• Risk of incorrect recognition of borders due to neighbouring structures Risk of incorrect recognition of borders due to neighbouring structures • Vertical borders close to the echogenic edges of the brainstem and Vertical borders close to the echogenic edges of the brainstem and choroid plexuschoroid plexus• Lateral borders rather narrow over the IT regionLateral borders rather narrow over the IT region• Square frame rather than rectangularSquare frame rather than rectangular• Smaller frameSmaller frame

IT vs NT semi-automated measurementIT vs NT semi-automated measurement

NTNT

ITIT

Moratalla J et al., UOG 2010

Results

Intraoperator SD (mm) 0.091

Median (IQR) (mm) 2.1 (1.9–2.4)

Operator 1

Manual

0.088

2.1 (1.8–2.3)

Operator 2

0.054

2.2 (1.9–2.4)

Operator 1

Semi-automated

0.067

2.2 (1.8-2.2)

Operator 2

Intraclass correlation 0.940 0.939 0.982 0.971

Mean difference (mm) –0.09 Reference* 0.01 –0.09

*The mean of the two manual measurements of operator 2 was considered the ‘gold standard’.

Results

*versus the gold standard (the mean of the two manual measurements of operator 2 ).

Good overall agreement with a trend towards larger measurements using the semi-automated approach

95% limits of agreement 95% limits of agreement

–0.26 and 0.23 mm

ManualManual

–0.17 and 0.17 mm

Semi-automatedSemi-automated

Interoperator & intermethod* –0.34 and 0.17 mm –0.20 and 0.23 mm

Intraoperator

Intra- and interoperator reliability of manual and semi-automated measurements of intracranial translucency

Karl et al., UOG 2012

Limitations

• Unknown reliability with variations in gestation and anatomy• Possible misdiagnosis of Blake’s pouch cyst as spina bifida

Strengths

• SonoNT can be used reliably for IT• Benefit for less experienced operators• High repeatability

Conclusion Manual IT measurements are reproducible

IT can be assessed reliably using the semi-automated NT algorithm, leading to standardization of the IT assessment process

Intra- and interoperator reliability of manual and semi-automated measurements of intracranial translucency

Karl et al., UOG 2012

Intra- and interoperator reliability of manual and semi-automated measurements of intracranial translucency

Karl et al., UOG 2012

Discussion points• What is the recurrence risk for spina bifida after having one affected

pregnancy?• What medical advice should be given to parents regarding a future

pregnancy in the above scenario?• If a recurrence occurs, will it always be limited to the fetal spine?• What is the difference between exencephaly, anencephaly and

encephalocele?• Can the above be diagnosed with confidence at 11–14 weeks’ gestation

by ultrasound?• Is measurement of intracranial translucency (IT) reproducible and

reliable enough to be tested in clinical practice?• Do we know how sensitive measurement of IT is in the detection of

spina bifida?