J Gandhara Med Dent Sci

October - December 2026

ORIGINAL ARTICLE

How to cite this article

DIAGNOSTIC ACCURACY OF ULTRASOUND IN DIAGNOSING CRANIOSYNOSTOSIS IN

INFANTS USING CT AS THE GOLD STANDARD

Kiran Zafar

, Naila Tamkeen

, Jawad Ali

, Khalid Kamal

, Sahar Gulab

, Faryal Bashir

Zafar K, Tamkeen N, Ali J, Kamal K,

Gulab S, Bashir F. Diagnostic

Accuracy Of Ultrasound in

Diagnosing Craniosynostosis In

Infants Using Ct as the Gold Standard.

J Gandhara Med Dent Sci.

2026;13(4):53-56.

Date of Submission: 06-03-2026

Date Revised: 15-09-2026

Date Acceptance: 15-09-2026

Trainee Medical Ocer, Department of

Radiology, Hayatabad Medical Complex

Specialist Registrar, Department of

Radiology, Hayatabad Medical Complex

Correspondence

Naila Tamkeen, Associate Professor

Department of Radiology, Hayatabad

Medical Complex

+92-334-9153074

drnaila814@gmail.com

ABSTRACT

OBJECTIVE

This study aimed to determine the diagnostic accuracy of cranial ultrasound

in detecting craniosynostosis in infants using computed tomography (CT) as

the reference standard.

METHODOLOGY

This cross-sectional validation study was conducted in the Department of

Radiology, MTI–Hayatabad Medical Complex, Peshawar. A total of 125

infants aged 2–6 months with clinical suspicion of craniosynostosis were

enrolled through non-probability consecutive sampling. All infants underwent

cranial ultrasound followed by CT scan. Ultrasound ndings were compared

with CT results, which were taken as the gold standard. Sensitivity,

specicity, positive predictive value (PPV), negative predictive value (NPV),

and overall diagnostic accuracy were calculated.

RESULTS

The mean age of infants was 4.02 ± 1.40 months. CT conrmed

craniosynostosis in 120 (96.0%) cases. Ultrasound demonstrated a sensitivity

of 79.17% (95% CI: 71.05–85.47%), specicity of 80.00% (95% CI: 37.55–

96.38%), PPV of 98.96% (95% CI: 94.33–99.82%), NPV of 13.79% (95% CI:

5.50–30.56%), and overall diagnostic accuracy of 79.20% (95% CI: 71.27–

85.39%).

CONCLUSION

Cranial ultrasound shows acceptable diagnostic accuracy with very high

PPV , supporting its use as a useful adjunct rst-line imaging tool for

craniosynostosis in settings with a high pre-test probability of disease,

particularly in resource-limited settings; however, its low NPV in this high -

prevalence referral cohort means a negative ultrasound cannot reliably

exclude the diagnosis, and CT remains essential for equivocal cases and

surgical planning.

KEYWORDS: Craniosynostosis; Cranial Ultrasound; Computed

Tomography; Diagnostic Accuracy; Infant Imaging; Radiation -Free

Screening; Pediatric Radiology

INTRODUCTION

Craniosynostosis is a congenital craniofacial anomaly

characterised by premature fusion of one or more

cranial sutures, resulting in abnormal skull morphology

and potential restriction of normal brain growth.

If not

identified and treated early, craniosynostosis may lead

to raised intracranial pressure, neurodevelopmental

delay, visual impairment, and psychosocial

consequences. Early diagnosis is therefore critical to

facilitate timely surgical intervention and optimise

neurodevelopmental outcomes.

Recent

epidemiological data estimate the global incidence of

craniosynostosis to be approximately 1 in 2,0002,500

live births, with an apparent rise in reported cases

attributed to improved diagnostic awareness and

imaging availability.

Craniosynostosis can occur as an

isolated, non-syndromic condition or as part of a

genetic syndrome. Non-syndromic craniosynostosis

accounts for nearly 75% of cases, commonly involving

sagittal, coronal, metopic, or lambdoid sutures. Genetic

mutations, often involving broblast growth factor

receptor pathways, contribute to approximately 20% of

cases, many arising de novo. Environmental factors

such as intrauterine constraint, maternal smoking, and

teratogen exposure have also been implicated.

Regardless of aetiology, early radiological conrmation

is essential once clinical suspicion arises.

Computed

tomography (CT) with three-dimensional reconstruction

remains the gold standard imaging modality for

diagnosing craniosynostosis due to its superior spatial

resolution and detailed depiction of cranial sutures. CT

also plays a vital role in surgical planning.

However,

CT exposes infants to ionising radiation, which has

been associated with an increased lifetime risk of

malignancy. Contemporary pediatric imaging

guidelines emphasise minimising radiation exposure

and adhering to the ALARA (As Low As Reasonably

http://doi.org/10.37762/jgmds.13-4.884

J Gandhara Med Dent Sci

October - December 2026

Achievable) principle, particularly in early infancy

Cranial ultrasound has emerged as an attractive

alternative imaging modality in this context.

It is non-

invasive, radiation-free, widely available, and cost-

effective. The presence of open fontanelles in early

infancy allows high-frequency linear probes to visualise

cranial sutures directly. Normal sutures appear as

hypoechoic gaps between hyperechoic bony margins,

whereas fused sutures demonstrate narrowing, bridging,

or complete absence of this gap.

Over the past ve

years, several studies have reported promising

diagnostic performance of ultrasound, particularly for

sagittal and coronal sutures, when performed by trained

radiologists using standardised protocols.

In low- and

middle-income countries such as Pakistan, the burden

of unnecessary CT scans is particularly signicant.

Limited access to advanced imaging, nancial

constraints, and lack of paediatric anaesthesia services

further complicate timely diagnosis.

Establishing

ultrasound as a reliable rst-line screening tool could

substantially reduce radiation exposure, healthcare

costs, and diagnostic delays. Despite growing

international evidence, local data on the diagnostic

accuracy of cranial ultrasound remain limited. This

study was therefore designed to evaluate the diagnostic

accuracy of cranial ultrasound in detecting

craniosynostosis in infants using CT scan as the

reference standard. By providing local evidence from a

tertiary care centre, this study aims to support safer,

more accessible diagnostic pathways and inform

imaging practices in resource-constrained settings.

METHODOLOGY

This cross-sectional validation study was conducted at

the Department of Radiology, MTI-Hayatabad Medical

Complex, Peshawar, after approval from the Hospital

Ethical Review Board (Approval No. 1725) and the

Research Evaluation Unit, CPSP Karachi. Written

informed consent was obtained from parents or legal

guardians. Infants aged 2-6 months referred with

clinical suspicion of craniosynostosis were enrolled

through non-probability consecutive sampling. Infants

outside the specied age range, those without consent,

and those who did not undergo both ultrasound and CT

during the same diagnostic work-up were excluded. As

recruitment was limited to clinically suspected cases at

a tertiary referral centre, the study population

represented a high-pretest-probability cohort. Cranial

ultrasound was performed before CT using a high-

frequency 14 MHz linear transducer. Major cranial

sutures were systematically examined in accessible

longitudinal and transverse planes. Patent sutures were

identified as hypoechoic gaps between echogenic bony

margins, while narrowing, loss of the normal gap, or

osseous bridging was considered suggestive of

premature fusion. Ultrasound was considered positive

when fusion of at least one cranial suture was detected.

CT of the skull served as the reference standard, with

suture patency assessed on bone-window images and

multiplanar or three-dimensional reconstructions where

available. Ultrasound interpretation preceded CT;

however, formal blinding of CT readers to ultrasound

findings was not prospectively documented. All 125

infants with paired ultrasound and CT results were

analysed. A formal a priori sample-size calculation was

not documented. The 120 CT-positive cases provided

adequate precision for estimating sensitivity, whereas

the small number of CT-negative cases limited the

precision of specicity and negative predictive value.

Sensitivity, specicity, positive and negative predictive

values, overall diagnostic accuracy, and disease

prevalence were calculated from a 22 table with 95%

exact binomial condence intervals. Data were

analysed using SPSS version 23. Suture-specic

diagnostic accuracy was not calculated because

complete paired suture-level ultrasound and CT data

were unavailable.

RESULTS

A total of 125 infants were included. The mean age was

4.02  1.40 months, with 61 (48.8%) males and 64

(51.2%) females (Table I). CT conrmed

craniosynostosis in 120 (96.0%) infants (Table II).

Ultrasound correctly identied 95 true-positive and 4

true-negative cases. There were 25 false-negative and 1

false-positive cases (Table III). Sensitivity, specicity,

PPV, NPV, and overall diagnostic accuracy were

79.17%, 80.00%, 98.96%, 13.79%, and 79.20%,

respectively (Table IV).

Table 1: Demographic Characteristics of Study Population

(n=125)

Variable Value

Mean age(months) 4.02  1.40

Gender (Male/Female) 61 (48.8%) / 64 (51.2%)

Total Infants 125

Table 2: CT-Conrmed Craniosynostosis (n=125)

CT Finding Frequency (%age)

Craniosynostosis Present 120 (96.0%; 95% CI 90.91-98.69)

Absent 5 (4.0%)

Table 3: Ultrasound Vs CT Diagnostic Comparison

CT Positive CT Negative

US Positive 95 1

US Negative 25 4

Diagnostic Accuracy of Ultrasound in Diagnosing Craniosynostosis

J Gandhara Med Dent Sci

October - December 2026

Table 4: Diagnostic Performance of Ultrasound

Parameter Estimate 95%

age

Sensitivity 79.17% 70.80-86.04

Specificity 80.00% 28.36-99.49

PPV 98.96% 94.33-99.97

NPV 13.79% 3.89-31.66

Overall Accuracy 79.20% 71.03-85.94

DISCUSSION

The present study demonstrates that cranial ultrasound

has acceptable diagnostic accuracy with a very high

positive predictive value for detecting craniosynostosis

in infants when compared with CT scan. These ndings

support the growing body of evidence advocating

ultrasound as a useful adjunct rst-line imaging tool in

early infancy for infants with a high pre-test probability

of craniosynostosis; however, given the low NPV

observed in this high-prevalence cohort, a negative

ultrasound nding alone should not be considered

sufcient to exclude the diagnosis when clinical

suspicion persists, and such cases should proceed to

CT. The sensitivity and specicity observed in this

study are consistent with recent international

literature.

Martino et al. (2022) reported sensitivity

exceeding 80% for ultrasound in detecting sagittal and

coronal suture fusion, while Najmi et al. (2023)

demonstrated comparable diagnostic accuracy in a

multicenter cohort. High PPV in our study indicates

that a positive ultrasound nding strongly correlates

with true craniosynostosis, allowing clinicians to

proceed with referral and management.

11,12

The low

negative predictive value observed in this study

warrants careful interpretation. This nding is primarily

attributable to the very high prevalence of CT-

confirmed craniosynostosis in our cohort, reecting

referral bias inherent to a tertiary care setting. Negative

predictive value is known to be highly inuenced by

disease prevalence rather than intrinsic test

performance. The 96% prevalence of CT-confirmed

craniosynostosis observed in this cohort is markedly

higher than the estimated population prevalence of

approximately 1 in 2,000-2,500 live births, and reects

the fact that infants were drawn exclusively from a

tertiary-care radiology referral pathway in which

clinical pre-selection for craniosynostosis was already

high, rather than from an unselected or primary-care

population.

By Bayes' theorem, NPV falls as pre-test

prevalence rises, even when the underlying sensitivity

and specicity of a test remain constant; the very low

NPV reported here is therefore a predictable

consequence of this referral-driven prevalence rather

than evidence of poor intrinsic test performance.

Consequently, these ndings are most directly

applicable to tertiary-referral populations with a

similarly high pre-test probability of disease, and

should not be extrapolated to lower-prevalence settings

such as primary-care or population-based screening,

where the same sensitivity and specicity would be

expected to yield a substantially higher NPV. Similar

trends have been reported in high-prevalence referral

populations.

False-negative ultrasound cases in our

study were mainly associated with partial suture fusion

and posterior suture involvement, particularly lambdoid

sutures. These limitations have been consistently

documented in prior studies and highlight the

importance of operator expertise and systematic

scanning protocols.

Suture-specific diagnostic

performance (i.e., separate sensitivity and specicity

for the sagittal, coronal, metopic, and lambdoid sutures)

is not reported in the present analysis. Consequently,

CT remains indispensable in cases with negative or

equivocal ultrasound ndings when clinical suspicion

persists, as well as for detailed preoperative planning.

Future multicenter studies should recruit a broader

spectrum of referred and non-referred infants, pre-

specify sample-size targets for both sensitivity and

specificity, use standardised ultrasound acquisition and

interpretation criteria, document reader experience and

blinding, and report suture-specific performance. Such

studies would better dene where ultrasound can safely

reduce CT use without delaying diagnosis.

CONCLUSIONS

Cranial ultrasound showed moderate sensitivity and a

high positive predictive value for craniosynostosis in

this highly selected tertiary-referral cohort. However,

the low negative predictive value and the wide

confidence interval around specicity mean that

ultrasound should not be used as a stand-alone rule-out

test. It may be used as a radiation-free initial adjunct or

triage examination in selected clinically suspected

infants, while CT remains the reference standard when

ultrasound is negative or equivocal despite persistent

clinical suspicion and for surgical planning.

LIMITATIONS

The single-centre tertiary-referral design, extreme

disease prevalence, the very small CT-negative

subgroup, and the resulting imprecision of specicity

and NPV. Operator experience was not stratified

analytically, CT-reader blinding to ultrasound was not

prospectively documented, and suture-specific paired

accuracy could not be assessed. Although the number

of CT-positive cases provides reasonable precision for

the sensitivity estimate, the study is underpowered for a

precise estimate of specicity.

CONFLICT OF INTEREST: None

FUNDING SOURCES: None

Diagnostic Accuracy of Ultrasound in Diagnosing Craniosynostosis

J Gandhara Med Dent Sci

October - December 2026

REFERENCES

1. Bautista G. Craniosynostosis: neonatal perspectives.

Neoreviews. 2021;22(4). doi:10.1542/neo.22-4-e250. PMID:

33795400.

2. Bashawieh OO, Alyami MH, Alghamdi MA, Alotaibi FE,

Alluhaybi AA, Altuhaini KS. Neurodevelopmental outcomes in

children with craniosynostosis: a retrospective cross-sectional

analysis. Childs Nerv Syst. 2026;42(1):49. doi:10.1007/s00381-

026-07128-9. PMID: 41604010.

3. Lewanda AF, Jabs EW, Sanchez-Lara PA. Craniosynostosis. In:

Pyeritz RE, Korf BR, Grody WW, editors. Emery and Rimoin's

Principles and Practice of Medical Genetics and Genomics:

Developmental Disorders. Elsevier; 2025. p. 81-104.

doi:10.1016/B978-0-12-815234-8.00004-5. PMID: 31335086.

4. Lattanzi W. Craniosynostosis: genetic basis, genes,

chromosomes, and resulting syndromes. In: Di Rocco C, Pang

D, Rutka JT, editors. Textbook of Pediatric Neurosurgery.

Cham: Springer; 2020. p. 1373-1391. doi:10.1007/978-3-319-

72168-2_56.

5. Russo C, Aliberti F, Ferrara UP, Russo C, De Gennaro DV,

Cristofano A, et al. Neuroimaging in non-syndromic

craniosynostosis: key concepts to unlock innovation.

Diagnostics (Basel). 2024;14(17):1842.

doi:10.3390/diagnostics14171842.

6. Aziz MM, Onyejesi C, Pyala R, Alattar O, Abdul AA,

Alagarswamy K, et al. Reducing radiation exposure in pediatric

CT imaging: strategies and alternatives in emergency

medicinea narrative review. J Emerg Crit Care Med. 2025;9.

doi:10.21037/jeccm-24-102.

7. Caro-Domínguez P, Lecacheux C, Hernandez-Herrera C,

Llorens-Salvador R. Cranial ultrasound for beginners. Transl

Pediatr. 2021;10(4):1117-1137. doi:10.21037/tp-20-399. PMID:

34012859.

8. Micovic M, Zivkovic B, Vukasinovic I, Jelovac D, Stojicic M,

Bascarevic V. Imaging modalities in craniosynostosis: a

systematic review and proposal of the ARCANA protocol for

multimodal radiation-free assessment. Diagnostics (Basel).

2025;15(20):2632. doi:10.3390/diagnostics15202632.

9. Massimi L, Bianchi F, Frassanito P, Calandrelli R, Tamburrini

G, Caldarelli M. Imaging in craniosynostosis: when and what?

Childs Nerv Syst. 2019;35(11):2055-2069. doi:10.1007/s00381-

019-04278-x. PMID: 31289853.

10. Proisy M, Bruneau B, Riaud L. How ultrasonography can

contribute to diagnosis of craniosynostosis. Neurochirurgie.

2019;65(5):228-231. doi:10.1016/j.neuchi.2019.09.019. PMID:

31586456.

11. Marino S, Ruggieri M, Marino L, Falsaperla R. Sutures

ultrasound: useful diagnostic screening for posterior

plagiocephaly. Childs Nerv Syst. 2021;37(12):3715-3720.

doi:10.1007/s00381-021-05324-3. PMID: 34453581.

12. Alizadeh H, Najmi N, Mehdizade M, Najmi N. Diagnostic

accuracy of ultrasonic examination in suspected

craniosynostosis among infants. Indian Pediatr. 2013;50(1):148-

150. doi:10.1007/s13312-013-0029-9. PMID: 23396787.

doi:10.3174/ajnr.A5063. PMID: 28183836.

13. Montoya JC, Eckel LJ, DeLone DR, Kotsenas AL, Diehn FE,

Yu L, et al. Low-dose CT for craniosynostosis: preserving

diagnostic benet with substantial radiation dose reduction.

AJNR Am J Neuroradiol. 2017;38(4):672-677.

14. Lyoo Y, Choi YH, Lee SB, Lee S, Cho YJ, Shin SM, et al.

Ultra-low-dose computed tomography with deep learning

reconstruction for craniosynostosis at radiation doses

comparable to skull radiographs: a pilot study. Pediatr Radiol.

2023;53(11):2260-2268. doi:10.1007/s00247-023-05717-3.

PMID: 37488451.

15. Kaasalainen T, Männistö V, Mäkelä T, Suojanen J, Nuorala A,

Heliövaara A, et al. Postoperative computed tomography

imaging of pediatric patients with craniosynostosis: radiation

dose and image quality comparison between multi-slice

computed tomography and O-arm cone-beam computed

tomography. Pediatr Radiol. 2023;53(8):1704-1712.

doi:10.1007/s00247-023-05644-3. PMID: 36967418.

16. Tao W, Goetti R. Evaluation of ultra-low-dose CT with tin lter

for craniosynostosis. J Med Imaging Radiat Oncol.

2025;69(1):28-34. doi:10.1111/1754-9485.13812. PMID:

39601235.

17. de Goederen R, Yang S, Pullens B, Wolvius EB, Joosten KFM,

Mathijssen IMJ. Evaluation of the OSA treatment protocol in

syndromic craniosynostosis during the rst 6 years of life. J

Plast Reconstr Aesthet Surg. 2021;74(10):2674-2682.

doi:10.1016/j.bjps.2021.03.033. PMID: 34112566.

18. Fahradyan A, Daneshgaran G, Homan TL, Wexler A, Francis

SH. Challenging the norm: is routine use of cranial CT in

evaluation of craniosynostosis necessary? J Craniofac Surg.

2021;32(7):2496-2499. doi:10.1097/SCS.0000000000007926.

PMID: 34705393.

LICENSE: JGMDS publishes its articles under a Creative Commons Attribution Non-Commercial Share-Alike license (CC-BY-NC-SA 4.0).

COPYRIGHTS: Authors retain the rights without any restrictions to freely download, print, share and disseminate the article for any lawful purpose.

It includes scholarlynetworks such as Research Gate, Google Scholar, LinkedIn, Academia.edu, Twitter, and other academic or professional networking sites.

Kiran Zafar

Concept & Design; Data Acquisition;

Data