Automatic recognition of the XLHED phenotype from facial images

Abstract

X-linked hypohidrotic ectodermal dysplasia (XLHED) is a genetic disorder that affects ectodermal structures and presents with a characteristic facial appearance. The ability of automated facial recognition technology to detect the phenotype from images was assessed . In Phase 1 of this study we examined if the age of male patients affected the technology's recognition. In Phase 2 we investigated how well the technology discriminated affected males cases from female carriers and from individuals with other ectodermal dysplasia syndromes. The system detected XLHED to be the most likely diagnosis in all genetically confirmed affected male patients of all ages, and in 55% of heterozygous females. Interestingly, patients with other ED syndromes were also detected by the XLHED-targeted analysis, consistent with shared developmental features. Thus the automated facial recognition system represents a promising non-invasive technology to screen patients at all ages for a possible diagnosis of ectodermal dysplasia, with greatest sensitivity and specificity for males affected with XLHED.

Studying Down syndrome recognition probabilities in Thai children with de‐identified computer‐aided facial analysis

“…In the research app, however, comparisons for controlled studies can be performed to evaluate quantifiable differences between facial gestalts of at least two groups. Binary and multiclass comparisons are made between the groups of photos using a cross validation scheme where all samples are randomly split in half multiple times, into training and test sets (Hadj‐Rabia et al, ; Liehr et al, ). Binary results are shown by a receiver operating characteristic (ROC) curve, respective area under the curve (AUC), and p value.…”

“…The original data set of patient photos used to train the DCNN were predominantly of Caucasian descent; therefore, its accuracy within other populations should be elucidated (Kruszka, Porras, Addissie, et al, ). Face2Gene's recognition ability was shown to have constantly improved by adding more photos of the same diseases among different ethnicities to the database (Hadj‐Rabia et al, ; Lumaka et al, ; Zarate et al, )…”

Reliability of prenatal detection of X‐linked hypohidrotic ectodermal dysplasia by tooth germ sonography

“…Mandibular hypoplasia is a consequence of oligodontia and was consistently observed prenatally as well as postnatally. Because of missing teeth, reduced alveolar processes, and mandibular hypoplasia, individuals with XLHED often have a typical facial appearance with a negative overjet . Absent or misshaped teeth are frequently the reason to seek medical help for individuals with previously undiagnosed XLHED and pose a significant psychosocial burden on the patients .…”

“…Because of missing teeth, reduced alveolar processes, and mandibular hypoplasia, individuals with XLHED often have a typical facial appearance with a negative overjet. 34,35 Absent or misshaped teeth are frequently the reason to seek medical help for individuals with previously undiagnosed XLHED 34 and pose a significant psychosocial burden on the patients. 36 Although oligodontia is not life-threatening, it affects the quality of life and may be one of probably multiple causes of impaired postnatal weight gain as observed in our cohort of children with XLHED.…”

Clinical versus automated assessments of morphological variants in twins with and without neurodevelopmental disorders

“…Instead, frontal facial photos taken by conventional cameras or cellular phones can be used (Basel‐Vanagaite et al, ; Gurovich et al, ). Face2Gene can be used by medical professionals and researchers who have limited access to trained dysmorphologists (Hadj‐Rabia et al, ) or in settings with limited socioeconomic resources (Lumaka et al, ).…”