APTD – My New Fetal Measurement using 2-D sonography

  • Can the anterior-posterior thigh diameter be used as an indicator for fetal age using two-dimensional sonography?

SAAD RAMZI ISMAIL; 

Please read my full paper , published with the JSDMS -USA. This research won the 3rd prize Canadian excellence in Sonography award -CHASMS,2005.FOR more information regarding this research please search google under my name Saad ramzi Ismail and then put APTD .My research was also published  in the RADIOGRAPHER JOURNAL -The South African Radiographer | ISSN 0258 0241 . 2005. This study was approved by the Ethics Committee of Charles Sturt University, Australia as my MSc  thesis project in Medical Ultrasound.

February 2008 Volume 14, Issue 1, Pages 45–51

Abstract:

This study evaluates the usefulness and direct correlation of a new method of predicting fetal age by the measurement of the anterior posterior high diameter (APTD) in normal 18 to 28 week pregnancies using two-dimensional sonography. Little published research exists in the area of fetal thigh biometry, specifically in the use of the APTD. The only study the author found was that of fetal thigh circumference. Continuing review of existing practices needs to be coupled with evaluation of alternate or additional methodology.

Material and methods:

This was a quantitative prospective study of 55 patients in a hospital in Alberta, Canada. APTDs were sonographically
measured. The normal range for each week of pregnancy was determined for reliability.

Results:

Significant correlation was found between the APTD and fetal age from simple line regression analysis, with 99.9% confidence intervals at each week from 18 to 28 weeks gestation. There was a correlation of 1 mm APTD per one week of fetal age. In addition R >0.93, P< than 0.001. The residual scatter plots confirmed the APTD validity.

Conclusion:

APTD is a reliable and valid method for assessing fetal age in a normal pregnancy and may be particularly useful when other parameters are unable to accurately predict fetal age. An accurate linear measurement of multiple fetal parameters allows a more complete profile of fetal growth and estimated date of delivery. APTD may also be useful in identifying fetal growth problems. All of the values of fetal age lie directly on the best-fit regression line. Since the coefficient of determination (Rsq) is very high, this model is very effective.

Can the anterior-posterior thigh diameter be used as an indicator for fetal age using two-dimensional sonography?

This study evaluates the usefulness and direct correlation of a new method of predicting fetal age by the measurement of the anterior posterior, thigh diameter (APTD) in normal 18 to 28 week pregnancies using two-dimensional sonography. Little published research exists in the area of fetal thigh biometry, specifically in the use of the APTD. The only study the author found was that of fetal thigh circumference. Continuing review of existing practices needs to be coupled with evaluation of alternate or additional methodology.

Material and methods: This was a quantitative prospective study of 55 patients in a hospital in Alberta, Canada. APTDs were sonographically measured. The normal range for each week of pregnancy was determined for reliability.

Results: Significant correlation was found between the APTD and fetal age from simple line regression analysis, with 99.9% confidence intervals at each week from 18 to 28 weeks gestation. There was a correlation of 1 mm APTD per one week of fetal age. In addition R >0.93, P< than 0.001. The residual scatter plots confirmed the APTD validity.

Conclusion: APTD is a reliable and valid method for assessing fetal age in a normal pregnancy and may be particularly useful when other parameters are unable to accurately predict fetal age. An accurate linear measurement of multiple fetal parameters allows a more complete profile of fetal growth and estimated date of delivery. APTD may also be useful in identifying fetal growth problems. All of the values of fetal age lie directly on the best-fit regression line. Since the coefficient of determination (Rsq) is very high, this model is very effective.

Keywords: APTD, parameters, weight estimation, biometry.

APRIL 2007 volume 45 number 1 THE SOUTH AFRICAN RADIOGRAPHER

www.sorsa.org.za

The participants were chosen by randomized selection of normal

pregnancies. All participants had an early (first trimester) sonogram at 12

week gestation to confirm the accuracy of the gestational age.

The author performed the study and compiled comparable data and

tables with the Hadlock et al tables for FL [31, 32]. The growth of the

APTD, outer to outer skin surface, was sonographically measured at the

middle point of the fetal femur in the sagital section and compared with

the fetal age from 18 to 28 weeks gestation. The second trimester period

was chosen because soft tissue accretion of the fetal thigh begins to

accelerate towards the end of this period.

The inclusion criteria for this study were: singleton uncomplicated

pregnancies with a normal fetus, and an informed consent form, read and

signed by the participants and approved by the hospital and Charles Sturt

University Ethics Committee.

The participants’ ages ranged between 18 and 35 years, with a mean

age of 26.5. The study population was of different ethnic groups in Alberta,

Canada, for example, Germans, natives Indians, Mennonites, Irish,

Hispanics, Ukrainians and east Indians.

Radiologists reported major congenital malformations, chromosomal

abnormalities, and maternal complications. The author did not release any

pathology information to the participants as they were informed to obtain

their reports from their physicians.

Routine transabdominal sonography was performed, including FL, BPD,

AC, and head circumference (HC). In addition, the author measured the

fetal APTD, from the middle point of the fetal femur in the sagittal section

of the fetal thigh using the femur length as a landmark. The APTD

measurements were analyzed and compared with fetal age using Hadlock

et al tables for femur length [31, 32].

Equipment used in this study was ATL 5000 and Philips Alegra 4500

(Bothell, WA). With 5 to 3 MHz transducers the fetal age was determined

by using measurements of the FL [31, 32].

A comparison was made between the APTD and the fetal age. The

diameter of the fetal thigh was measured in the same portion of fetal thigh

every time by measuring the mid-point of the femur. Eleven groups were

studied. Each group comprised five participants who were all in the same

pregnancy period, namely from 18 to 28 weeks.

Protocol used in this study

Starting with transducer at the fetal abdominal circumference.

1. Transducer was moved inferiorly to transect the fetal bladder.

2. Transducer was rotated 30 degrees to view the fetal femur.

3. Transducer was rotated until a sagittal view of the fetal thigh was

obtained (Figure 1).

4. Distal femoral epiphyses which are usually present after 32 weeks

gestation were excluded.

5. The fetal knee was identified.

6. In cases when a double line was seen in the fetal thigh the inner

line was measured or the scan was repeated until a smooth

(sagittal) line of the fetal thigh was obtained (Figure 2). Note that

this double line can be corrected by obtaining a perfect sagittal

view of the fetal thigh. Otherwise the curve of the thigh adds extra

false line to the real outer skin surface of the fetal thigh in the

lateral or medial section. The thigh is convex in the anterior part

and concave in the posterior part so geometrically we are dealing

with a cylinder and not a flat surface.

7. Real-time sonographic equipment was used with 3.0, 3.5, and 5.0

MHz transducers frequencies to obtain the images. Note that

posterior shadowing from the femur bone is limited and will not

affect the over all measurements.

8. Freeze-frame and electronic calipers were used as they are more

sensitive tools to provide accurate measurements of the fetal

thigh.

9. Zoom capability was used to outline the fetal thigh, namely the

outer skin surface which allowed the author to see the skin surface

better therefore the chances of errors were reduced.

10. Hadlock et al table for FL [31, 32] was used to compare with

APTD

Measurements

The FL of the fetus was scanned in each of the pregnant participants as a

sagittal view (Figures 1 and 2).

Figure 1. Label A is showing the wrong way to measure fetal thigh (coronal)

and Label B is showing the correct way to measure the anterior-posterior

thigh diameter (APTD) in sagittal plane (profile).

Figure 3. The sagittal section of the fetal thigh is showing the measurement

of the femur length. The arrow is showing the fetal knee.

Magnification can be a helpful tool.

Figure 2. The white arrow is showing the double line of the fetal thigh. The

correct measurement of the anterior-posterior thigh diameter would be the

second line marked by the number (1) arrow in the real anterior wall of the

fetal thigh; as this is the true skin line. The second line marked by number (2)

arrow is part of the thigh tissue as the sound waves travels through the

convex area can be corrected by scanning in a good sagittal plane.

Figure 4. Sagittal plane of the fetal thigh is showing the femur length with

one of the calipers in the mid point of the femur length.

THE SOUTH AFRICAN RADIOGRAPHER volume 45 number 1 APRIL 2007

www.sorsa.org.za

16

1. The femur length was measured using callipers, namely first

caliper placed in exact middle point of the fetal femur; for example,

if the femur length was 2.4 cm then the first calliper was moved

until the measurement read 1.2 cm (Figures 3 and 4).

2. The first calliper was carefully moved to the outer surface of the

fetal anterior thigh (Figure 5); the skin surface was measured and

not the extra double line created by the sound waves that travel

through the convex part of the thigh in para-sagittal planes. Note

that scanning the fetal thigh in the sagittal plane can make

correction and smooth the skin surface of the fetal thigh. The

second calliper was moved to the outer surface of the posterior

thigh, and entered (Figure 6).

Calculations

Each one millimetre of the APTD, or the posterior-anterior thigh diameter

(PATD) measurements was taken to be equal to one-week. For example,

1.90 cm will be equal to 19 weeks gestation, and 2.80 cm will be equal to

28 weeks gestation. Multiply 1.428 with any fraction of a millimetre. This

number 1.428 was obtained from 10 mm divided by 7 days, for example,

APTD of 2.68 cm (26.8 mm) calculates to 26 weeks plus 0.8 x 1.428 =

0.1424 day. This is added to the 26 weeks equalling 27.0 weeks and 1.4

day. The APTD was found to be relatively constant, one mm equal to one

week.

Serial measurements should however be obtained. The measurements

should be repeated with zooming capability and electronic callipers; the

serial measurements range should be less than 1 mm. If these

measurements do not match the fetal age obtained by using Hadlock et al

tables for FL [31, 32] a follow-up scan should be recommended.

Results

Measurements of fetal femur lengths of the 55 pregnant participants who

met the criteria were correlated with the APTD and used to construct

tables and graphs. There was significant correlation between the APTD and

fetal age. Using a simple linear regression for this study, more than 99.9

% confidence intervals were found at each week of the eleven groups from

18 to 28 weeks gestation (R > 0.93), and (p less than 0.001). The APTD

was positively correlated with fetal age (Table 1 and Graphs 1, 2, 3 and 4).

Eleven gestational periods from 18 to 28 weeks were analyzed, each

period included five different measurements of the femur lengths

compared to the fetal age and to the ATPD with mean +/-2SD. Femur

length measured from 2.70 cm to 5.50 cm over all gestational periods, the

mean being 4.3. Fetal weight ranged between 310 grams and 1400

grams, the mean being 629 grams. The APTD ranged between 1.80 to

2.87 cm, with the mean at 2.36 cm. Linear growth was obtained in each

gestational period from 18 to 28 weeks, and compared with the Hadlock et

al tables [31, 32]. In addition a linear growth of fetal weight was observed

in the graphs (see Graph 1). The ATPD when converted to millimetres and

compared with the fetal age, was found to be a consistent and valid

measurement by using the scatter plots (Graphs 2 and 3). The standard

error of estimates using the APTD was significantly lower at .08664 than

that using femur length at .2436. The variability estimates from Hadlock et

al table [31] for femur length versus fetal age from 18 to 30 weeks have

indicated ±1.8 weeks to 2.4 weeks. The APTD table in this study shows

±3 days variability (see Table 1). Adjusted R square variance was >.99 for

both models.

Statistical analysis

Regression: APTD (cm) and fetal age (weeks)

The standard error of estimation is very low at .08664. This indicates a

strong good fit of this model. The spread of values for the dependent

variable, fetal age, around the mean value of the independent variable is

very narrow. About 70% of the values of fetal age will lie ± .08664 from

the mean of APTD using a nova method.

Discussion

The simplicity of the application found by this study is really its greatest

advantage. The idea is new as the only study done previously was the fetal

Figure 5. The first caliper is moved to the real outer skin of the anterior wall

of the fetal thigh.

Figure 6. The second caliper is moved to the posterior wall of the fetal thigh.

Enter and log the measurement of the anterior posterior

thigh diameter (APTD).

Graph 1: Regression Line for APTD

All of the values of fetal age lie directly on the “Best Fit” Regression Line.

Graph 2: Regression Line for Femur Length and Gestational age.67

Some of the values of fetal age in Dr Hadlock’s model lie

slightly above or slightly below the “Best Fit” Regression Line.

R Square was .994 and SEE was .24362.

thigh circumference and not the APTD. Accuracy of fetal age, weight, and

estimated delivery date should be improved if multiple predictors are used,

especially when it is difficult to obtain fetal head biometry; for example,

when the head is too low in the pelvis, hydrocephalus, anencephaly, and

fetal renal disease.

New methods for estimating fetal body weight and fetal age without

head measurement are therefore required. Reliable new methods of fetal

biometry could be very beneficial in reducing overall fetal biometry errors

and increasing the reliability of fetal biometry. Results of this study show

that the ATPD predicts second trimester growth with high validity and

reliability.

The very simple correlation in this study of 1 mm APTD per week of

fetal age is new and useful information. Measuring thigh parameter can be

a convenient method for determining fetal growth in the second trimester.

The APTD may have a role in quality control of second trimester sonogram

examination and may help in the diagnosis of fetal growth abnormalities.

The APTD may be used as an indicator of fetal biometric disturbance

enabling the physician to better manage the pregnancy. Diabetes mellitus

is one cause of intrauterine growth restriction (IUGR), and may affect the

femur length [10,23,27]. Diabetes mellitus may also affect the fetal body

mass and consequently the abdominal circumference and fetal thigh [28]

hence the APTD can be used not only as indictor for fetal age but also to

detect IUGR. Renal pathology, such as hydronephrosis or congenital renal

malformation, can affect the fetal abdominal circumference, making this

measurement unreliable as an indicator of fetal age. Using combined

parameters may be superior to the use of each measurement alone as a

Table I. Correlation between anterior-posterior thigh diameter (APTD-CM) and fetal age (GA-WK) 50th percentile values for fetal femur length

are shown below, (n=55).

Femur Length

(cm) From

Hadlock Table67, 68

Fetal age (wk)

using Hadlock

Graph 3: the residual (error), scatter plot and validity of the APTD

Normally distributed residuals scatter plot

Graph 4: Residual (error) scatter plot – femur length from Dr Hadlock’s Table

67. Pattern of residuals show curvilinear relationship with regression line

marker of trisomy 21 [30]. In addition, it can be difficult in practice to

obtain a good fetal thigh circumference, or fetal hands, feet and ears to

obtain fetal biometry. This study shows that the fetal APTD provides a more

accurate linear measurement of the fetus, thus generating a more

complete profile of the fetus.

Significant correlations of APTD with fetal age indicate that this is a

reliable method and is particularly useful when other fetal parameters may

not accurately predict fetal age or if they are difficult to obtain. If the age

predicted from the APTD does not match the age using the femur length,

other factors, such as intrauterine growth restriction or maternal and fetal

nutrition deficits, should be considered.

The soft tissue accretion of the fetal thigh also depends on the

generalized nutritional status of the fetus but such increase in the soft

tissue is usually more marked after the 30th week of gestation. The APTD

measurements obtained from the 11 groups correlated well with the fetal

age and repeated five times for each gestational group between 18 and 28

weeks.

Tables for femur length versus gestational age from 18 to 30 weeks

were ± 1.8 to ± 2.4 weeks while the variability estimates in the APTD

table was ± 3days. Both models predict the fetal age very well, but

compared to FL using APTD produces a model with better goodness of fit

based on differences in the standard error of estimates (SEE) between the

two of them, and on interpreting the best fit regression lines for both

models. The spread of values for the dependent variable is narrower

around the mean of the independent variable in the APTD model and wider

in the FL model.

The SEE of .2436 obtained for FL versus gestational age is higher than

that obtained in the analysis with APTD. This indicates a weaker goodness

of fit of this model. The spread of values for the dependent variable around

the mean value of the independent variable is wider; 68% of the values of

fetal age will lie ± .2436 from the mean of APTD. Model statistics (F, t, and

standardized Beta) are significant for both models. Beta (APTD) = 10.0

(SEE = . 037), Beta (FL) = 3.79 (SE = . 039). T =273.07 for GA x APTD

Model = 96.87 for GA x FL Model

Conclusion

APTD was found to be a valid and reliable index for estimating fetal age.

Further research to study the relationship between APTD versus fetal

weight and IUGR is needed. More research is also needed to study the

APTD measurements from 12 to 40 weeks gestational age and in a larger

population in order to be more statistically significant.

Declaration

No financial support was granted to this research and no commercial

affiliation was involved.

Acknowledgements

The author wishes to express his deepest appreciation to his project

supervisors Dr Karen Hofmann and Karen Pollard for their guidance and

advice. In addition thank you to DR Philip Hughes, S Desilva, A Zammit, Gary

Kachure, Dr K Game, Dr V Botha, and Dr S Benade for their assistance.

Grateful thank you to Ms Jean Spitz, Valletta Lawrence, and Siti Arabiah.

This research is dedicated to the memories of my parents and brothers.

References

1. Jensen SL, et al: Fetal age assessment based on sonogram head

biometry and the effect of maternal and fetal factors. Acta obstetricia

and gynecologica Scandianvica; 2004; 83: Issue 8, P716.

2. Widdowson E.M, Mcance R.A: The determination of growth and form.

Proc. Royal Soc. Lond.1974; 185:1-17.

3. Barker D.J.P: The fetal origins of adult disease. Proc. Royal Soc.

Lond.1994a; 262:37-43.

4. Barker D.J.P: Mothers, babies, and disease in later life. B.M.J

Publishing. London. 1994b.

5. Kurmanvicius J: Ultrasonographic fetal weight estimation: accuracy

of formulas and accuracy of examiners by birth weight from 500 to

5000 grams. J Perinat Med 2004; 32(2004) 155-161.

6. Cheng Mc, et al: Birth weight distribution of Singapore, Chinese,

Malay, and Indian infants from 34 to 42 weeks gestation. J Obstet

Gynaecol, Br Common W 1972; V (2): 149-53.

7. Freman MG, et al: Indigent black and Caucasian birth weight-fetal

age tables. Pediatrics 1970 Jul; 46(1): 9-15.

8. Goldenberg RL, et al: Black-white differences in newborn

anthropometrics measurements. Obstet Gynecol, 1991 Nov. 78 (5 pt

1): 782-8.

9. Jansen GM, et al: The effect of high altitude and other risk factors on

birth weight: independent or interactive effects. A.M. J Public Health,

1997, 87- (6): 1003-7.

10. Berkus MD, et al: Glucose tolerance tests: degree of glucose

abnormality correlates with neonatal outcome. Obstet Gynecol, 1993;

81(3): 344- 8.

11. Yoshida S, Nuno N, and Kageawa H: Prenatal detection of high-risk

group for intrauterine growth restriction based on sonographic fetal

biometry. Int J Gynaecol Obstet 2000; 68 (3): 225-32

12. Ratanasiri T, et al: Comparison of the accuracy of ultrasonic fetal

weight estimation by using the various equations. J Med Assoc Thai.

2002 Sept 85 (9): 962-7.

13. Vintzileos AM, Neckles S, Campbell WA, et al: Sonogram fetal thigh

calf circumference and fetal age-independent fetal ratios in normal

pregnancy: Journal of sonogram in medicine, 1985; Vol 4, Issue 6,

287-292.

14. Jeanty. P, et al: Estimation of fetal age from measurement of fetal

long bones. Journal of sonogram in medicine, 1984; Vol 3, Issue 2,

P75-79, and. Jeanty P; et al: Fetal Limb Volume; Anew parameter to

assess fetal growth and nutrition Sonogram Med 1985; 4:273-282.

15. Bunduki V, et al: Fetal nasal bone length: reference range and

clinical application in sonogram screening for trisomy 21: sonogram in

Obstetric and Gynaecology, Feb 2003, Vol 21, no 2, pp 160 (5).

16. Campbell J, Henderson A, Campbell S: The fetal femur / foot length

ratio: a new parameter to assess dysplastic limp reduction: Obs / Gyn,

1988; 72:181-184.

17. Streeter GL: Weight sitting height, head size, foot length, and

menstrual age for the human embryo, Taylor & Francis, Annals of

Human Biology.1920; 55: P 156.

18. Chatterg H S, Adhate A: Ultrasonic imaging of fetal foot for

evaluation of fetal age: proceeding of the WFUMP 1986; July, 206: 14

– 19.

19. Goldstien A. Tamir E. Z, Itskovitz J: Growth of fetal orbit and lens in

normal pregnancies: sonogram in: Obs / Gyn.1998; Vol 12, issue 3, p

175.

20. Konje J. C, Abrams K. R, Bell S.C & Taylor D.J: Determination of

fetal age after 24th weeks of gestation, from fetal kidney length

measurements; Sonograms in, Obs / Gyn 2002; 19(6), 592-597.

21. Mercer BM, et al: Fetal foot length as a predictor of fetal age; AMJ

Gynecol.1987; 156(2): 350-5.

22. Chen CP, et al: Prediction of fetal macrosomia by single ultrasonic

fetal biometry. J Formosa Med Assoc. 1993 Jan; 92(1): 24-8.

23. Flanagan DE, et al: Fetal growth and the physiological control of

glucose tolerance in adults: a minimal model analysis AMJ, Physiology

endocrinal Metab. 2002; 278(4).

24. Robinson H.P, et al: A critical evolution of sonar crown-rump length

measurements. BJOG, 1975; 82: 702-710.

25. Evans J.A, et al: BMUS, Ultrasonic fetal measurement survey, part 4,

BMUS Bulletin, 1989; 52; 4-10

26. Deter R.L, et al: Fetal head and abdominal circumferences: A critical

re-evaluation of the relationship to menstrual age. Journal of clinical

sonogram, 1982; 10: 365-372.

27. Benson C.B, et al: Intrauterine growth retardation: Predictive value of

sonogram criteria for antenatal diagnosis. Radiology, 1986; 160: 415-

417.

28. Gillman MW, et al: Maternal gestational diabetes, birth weight, and

adolescent obesity. Paediatrics. 2003: 111(3): 221-6.

29. Degani S: Fetal biometry: Clinical, pathological, and technical

consideration. Obstetric Gynaecology Survey. 2001, 56(3): P 159-67.

30. Thomas C. et al: Cerebella and frontal lobe hypoplasia in fetuses

with trisomy 21 usefulness as combined Ultrasound markers. RSNA,

Radiology. 2000; 214: 533-538.

31. Hadlock FP, Deter RL, Harrist RB, &Park SK: Fetal femur length as a

predictor of estimating fetal age: Computer-assisted analysis of

multiple fetal growth parameters. Radiology. 1984; 152:497-501.

32. Hadlock FP, Deter RL, Harrist RB, & Park SK: Fetal femur length as a

predictor of menstrual age: Sonographically measured. AJR, 1982;

February 2008Volume 14, Issue 1, Pages 45–51

 Can the anterior–posterior thigh diameter be used as an indicator for fetal age using two-dimensional sonography?

Saad Ramzi Ismail, PhD. Calgary- Alberta

Abstract

This study evaluated the usefulness and direct correlation of a simple new method of predicting fetal age by measurement of the anterior–posterior thigh diameter (APTD) in a normal 18 to 28 week pregnancies using two-dimensional sonography. Little published research exists in the area of fetal thigh biometry, specifically in the use of the anterior–posterior fetal thigh diameter (APTD). The only study I found was that of fetal thigh circumference. Continuing review of existing practices needs to be coupled with evaluation of alternate or additional methodology.

Materials and methods

This was a quantitative prospective study of 55 patients in High Level General Hospital, Alberta, Canada. Anterior–posterior thigh diameters (APTD) were sonographically measured. The normal range for each week of pregnancy was determined five times for reliability.

Results

Significant correlation was found between (APTD) and fetal age from simple line regression analysis, with 99.993% confidence intervals at each week from 18 to 28 weeks gestation. There was a correlation of 1 mm APTD per 1 week of fetal age. In addition R > 0.93, P < 0.001. The residual scatter plots confirmed the APTD validity.

Conclusion

APTD is a reliable and valid method for assessing fetal age in a normal pregnancy and may be particularly useful when other parameters are unable to accurately predict fetal age. An accurate linear measurement of multiple fetal parameters allows a more complete profile of fetal growth and estimated date of delivery (EDD). APTD may also be useful in identifying fetal growth problems. All of the values of fetal age lie directly on the “best-fit” regression line. Since the coefficient of determination (Rsq) is very high, this model is very effective.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top