ESPE Abstracts (2021) 94 P1-129

ESPE2021 ePoster Category 1 Growth A (10 abstracts)

Computer-aided facial analysis as a tool to identify patients with Silver-Russell syndrome and Prader-Willi syndrome

Silvia Ciancia 1,2 , Wesley J. Goedegebuure 1 , Lionne N. Grootjen 1 , Anita C.S. Hokken-Koelega 1 , Gerthe F. Kerkhof 1 & Daniëlle C. van der Kaay 1

1Department of Pediatrics, Subdivision of Endocrinology, Erasmus University Medical Center-Sophia Children’s Hospital, Rotterdam, Netherlands; 2Post-Graduate School of Pediatrics, Department of Medical and Surgical Sciences for Mother, Children and Adults, University of Modena and Reggio Emilia, Modena, Italy

Introduction: Genetic syndromes often show suggestive facial features that provide clues for the diagnosis. Considering the high number of genetic syndromes and the possible overlap of some features, memorizing facial gestalt is a challenging task for clinicians. DeepGestalt technology, and its app Face2Gene, has a growing impact on the diagnosis and management of genetic diseases by analyzing the features detected in one or more facial images of affected individuals.

Methods: 23 pediatric patients with clinically or genetically diagnosed Silver-Russell syndrome (SRS) and 29 pediatric patients with genetically confirmed Prader-Willi syndrome (PWS) were enrolled between December 2020 and April 2021. One frontal picture of each patient was acquired. Top-1, top-5 and top-10 sensitivity was analyzed. Correlation with the specific genetic diagnosis was investigated. When available, pictures of the same patient at different ages were compared.

Results: In the SRS cohort Face2Gene showed a top-1, top-5 and top-10 sensitivity of 39%, 65% and 91% respectively. Two patients (8.7%) were not correctly diagnosed by the app; this did not change after adding clinical features. Thirteen SRS patients had SRS suggested with high level of probability. In 44% of genetically confirmed patients SRS was the first syndrome suggested while in clinically diagnosed patients, SRS was suggested as top-1 in 28% of cases (P = 0.49). In 22% of SRS patients a picture at the age of the diagnosis or at a younger age than the age of enrollment in the study was available. For pictures obtained at a younger age (mean age at diagnosis 2.61 years, range 0.17-9 years; mean age at enrollment 7.3 years, range 1-17 years), Face2Gene performed better in younger patients and for all patients SRS was suggested as top-1, albeit with variable degree of probability. In the PWS cohort the top-1, top-5 and top-10 sensitivity were 76%, 97% and 100% respectively. Twenty-one percent of PWS patients had PWS suggested with high level of probability. PWS was suggested as top-1 in 60% of patients genetically diagnosed with mUPD and in 83% of patients presenting with paternal deletion of chromosome 15q11-13.

Conclusions: Face2Gene app can be a useful tool to support clinicians in the diagnosis of SRS and PWS. The sensitivity is higher in PWS patients and was comparable throughout all age ranges. In SRS patients, the app performed better in the younger age group, probably due to more pronounced facial features at a younger age.

Volume 94

59th Annual ESPE (ESPE 2021 Online)

22 Sep 2021 - 26 Sep 2021

European Society for Paediatric Endocrinology 

Browse other volumes

Article tools

My recent searches

No recent searches.