Facial dysmorphism is influenced by ethnic background of the patient and of the evaluator

Increasing use of artificial intelligence in genomic medicine for cancer care- the promise and potential pitfalls

The field of genomic medicine produces large datasets, which need to be rapidly analysed to produce clinically actionable insights in cancer care. Artificial intelligence thrives on data, processing and learning from datasets with a degree of accuracy and efficiency that traditional computing algorithms can not achieve. Based on a patient's genome sequence, AI could allow earlier detection of cancer, inform personalised treatment plans and provide insights into prognostication. However, this valuable tool is met with skepticism, with stakeholders concerned over data security, liability for AI's mistakes due to hallucination and the threat to clinical jobs. This review highlights both the benefits and potential problems of using AI in genomic medicine for cancer care, with the aim to lessen the knowledge gap between clinicians and data scientists and facilitate the future deployment of AI in cancer care.

Mitigating Algorithmic Bias in AI-Driven Cardiovascular Imaging for Fairer Diagnostics

Background/Objectives: The research addresses algorithmic bias in deep learning models for cardiovascular risk prediction, focusing on fairness across demographic and socioeconomic groups to mitigate health disparities. It integrates fairness-aware algorithms, susceptible carrier-infected-recovered (SCIR) models, and interpretability frameworks to combine fairness with actionable AI insights supported by robust segmentation and classification metrics. Methods: The research utilised quantitative 3D/4D heart magnetic resonance imaging and tabular datasets from the Cardiac Atlas Project's (CAP) open challenges to explore AI-driven methodologies for mitigating algorithmic bias in cardiac imaging. The SCIR model, known for its robustness, was adapted with the Capuchin algorithm, adversarial debiasing, Fairlearn, and post-processing with equalised odds. The robustness of the SCIR model was further demonstrated in the fairness evaluation metrics, which included demographic parity, equal opportunity difference (0.037), equalised odds difference (0.026), disparate impact (1.081), and Theil Index (0.249). For interpretability, YOLOv5, Mask R-CNN, and ResNet18 were implemented with LIME and SHAP. Bias mitigation improved disparate impact (0.80 to 0.95), reduced equal opportunity difference (0.20 to 0.05), and decreased false favourable rates for males (0.0059 to 0.0033) and females (0.0096 to 0.0064) through balanced probability adjustment. Results: The SCIR model outperformed the SIR model (recovery rate: 1.38 vs 0.83) with a -10% transmission bias impact. Parameters (β=0.5, δ=0.2, γ=0.15) reduced susceptible counts to 2.53×10-12 and increased recovered counts to 9.98 by t=50. YOLOv5 achieved high Intersection over Union (IoU) scores (94.8%, 93.7%, 80.6% for normal, severe, and abnormal cases). Mask R-CNN showed 82.5% peak confidence, while ResNet demonstrated a 10.4% accuracy drop under noise. Performance metrics (IoU: 0.91-0.96, Dice: 0.941-0.980, Kappa: 0.95) highlighted strong predictive accuracy and reliability. Conclusions: The findings validate the effectiveness of fairness-aware algorithms in addressing cardiovascular predictive model biases. The integration of fairness and explainable AI not only promotes equitable diagnostic precision but also significantly reduces diagnostic disparities across vulnerable populations. This reduction in disparities is a key outcome of the research, enhancing clinical trust in AI-driven systems. The promising results of this study pave the way for future work that will explore scalability in real-world clinical settings and address limitations such as computational complexity in large-scale data processing.

Toward 3D facial analysis for recognizing Mendelian causes of autism spectrum disorder

Recognizing Mendelian causes is crucial in molecular diagnostics and counseling for patients with autism spectrum disorder (ASD). We explored facial dysmorphism and facial asymmetry in relation to genetic causes in ASD patients and studied the potential of objective facial phenotyping in discriminating between Mendelian and multifactorial ASD. In a cohort of 152 ASD patients, 3D facial images were used to calculate three metrics: a computational dysmorphism score, a computational asymmetry score, and an expert dysmorphism score. High scores for each of the three metrics were associated with Mendelian causes of ASD. The computational dysmorphism score showed a significant correlation with the average expert dysmorphism score. However, in some patients, different dysmorphism aspects were captured making the metrics potentially complementary. The computational dysmorphism and asymmetry scores both enhanced the individual expert dysmorphism scores in differentiating Mendelian from non-Mendelian cases. Furthermore, the computational asymmetry score enhanced the average expert opinion in predicting a Mendelian cause. By design, our study does not allow to draw conclusions on the actual point-of-care use of 3D facial analysis. Nevertheless, 3D morphometric analysis is promising for developing clinical dysmorphology applications in diagnostics and training.

GestaltMatcher Database - A global reference for facial phenotypic variability in rare human diseases

The most important factor that complicates the work of dysmorphologists is the significant phenotypic variability of the human face. Next-Generation Phenotyping (NGP) tools that assist clinicians with recognizing characteristic syndromic patterns are particularly challenged when confronted with patients from populations different from their training data. To that end, we systematically analyzed the impact of genetic ancestry on facial dysmorphism. For that purpose, we established the GestaltMatcher Database (GMDB) as a reference dataset for medical images of patients with rare genetic disorders from around the world. We collected 10,980 frontal facial images - more than a quarter previously unpublished - from 8,346 patients, representing 581 rare disorders. Although the predominant ancestry is still European (67%), data from underrepresented populations have been increased considerably via global collaborations (19% Asian and 7% African). This includes previously unpublished reports for more than 40% of the African patients. The NGP analysis on this diverse dataset revealed characteristic performance differences depending on the composition of training and test sets corresponding to genetic relatedness. For clinical use of NGP, incorporating non-European patients resulted in a profound enhancement of GestaltMatcher performance. The top-5 accuracy rate increased by +11.29%. Importantly, this improvement in delineating the correct disorder from a facial portrait was achieved without decreasing the performance on European patients. By design, GMDB complies with the FAIR principles by rendering the curated medical data findable, accessible, interoperable, and reusable. This means GMDB can also serve as data for training and benchmarking. In summary, our study on facial dysmorphism on a global sample revealed a considerable cross ancestral phenotypic variability confounding NGP that should be counteracted by international efforts for increasing data diversity. GMDB will serve as a vital reference database for clinicians and a transparent training set for advancing NGP technology.

Advancing Fairness in Cardiac Care: Strategies for Mitigating Bias in Artificial Intelligence Models Within Cardiology

In the dynamic field of medical artificial intelligence (AI), cardiology stands out as a key area for its technological advancements and clinical application. In this review we explore the complex issue of data bias, specifically addressing those encountered during the development and implementation of AI tools in cardiology. We dissect the origins and effects of these biases, which challenge their reliability and widespread applicability in health care. Using a case study, we highlight the complexities involved in addressing these biases from a clinical viewpoint. The goal of this review is to equip researchers and clinicians with the practical knowledge needed to identify, understand, and mitigate these biases, advocating for the creation of AI solutions that are not just technologically sound, but also fair and effective for all patients.

Loss-of-Function Variants in CUL3 Cause a Syndromic Neurodevelopmental Disorder

OBJECTIVE

De novo variants in cullin-3 ubiquitin ligase (CUL3) have been strongly associated with neurodevelopmental disorders (NDDs), but no large case series have been reported so far. Here, we aimed to collect sporadic cases carrying rare variants in CUL3, describe the genotype-phenotype correlation, and investigate the underlying pathogenic mechanism.

METHODS

Genetic data and detailed clinical records were collected via multicenter collaboration. Dysmorphic facial features were analyzed using GestaltMatcher. Variant effects on CUL3 protein stability were assessed using patient-derived T-cells.

RESULTS

We assembled a cohort of 37 individuals with heterozygous CUL3 variants presenting a syndromic NDD characterized by intellectual disability with or without autistic features. Of these, 35 have loss-of-function (LoF) and 2 have missense variants. CUL3 LoF variants in patients may affect protein stability leading to perturbations in protein homeostasis, as evidenced by decreased ubiquitin-protein conjugates in vitro. Notably, we show that 4E-BP1 (EIF4EBP1), a prominent substrate of CUL3, fails to be targeted for proteasomal degradation in patient-derived cells.

INTERPRETATION

Our study further refines the clinical and mutational spectrum of CUL3-associated NDDs, expands the spectrum of cullin RING E3 ligase-associated neuropsychiatric disorders, and suggests haploinsufficiency via LoF variants is the predominant pathogenic mechanism. ANN NEUROL 2024.

The impact and future of artificial intelligence in medical genetics and molecular medicine: an ongoing revolution

Artificial intelligence (AI) platforms have emerged as pivotal tools in genetics and molecular medicine, as in many other fields. The growth in patient data, identification of new diseases and phenotypes, discovery of new intracellular pathways, availability of greater sets of omics data, and the need to continuously analyse them have led to the development of new AI platforms. AI continues to weave its way into the fabric of genetics with the potential to unlock new discoveries and enhance patient care. This technology is setting the stage for breakthroughs across various domains, including dysmorphology, rare hereditary diseases, cancers, clinical microbiomics, the investigation of zoonotic diseases, omics studies in all medical disciplines. AI's role in facilitating a deeper understanding of these areas heralds a new era of personalised medicine, where treatments and diagnoses are tailored to the individual's molecular features, offering a more precise approach to combating genetic or acquired disorders. The significance of these AI platforms is growing as they assist healthcare professionals in the diagnostic and treatment processes, marking a pivotal shift towards more informed, efficient, and effective medical practice. In this review, we will explore the range of AI tools available and show how they have become vital in various sectors of genomic research supporting clinical decisions.

Expanded phenotypic spectrum of neurodevelopmental and neurodegenerative disorder Bryant-Li-Bhoj syndrome with 38 additional individuals

Bryant-Li-Bhoj syndrome (BLBS), which became OMIM-classified in 2022 (OMIM: 619720, 619721), is caused by germline variants in the two genes that encode histone H3.3 (H3-3A/H3F3A and H3-3B/H3F3B) [1-4]. This syndrome is characterized by developmental delay/intellectual disability, craniofacial anomalies, hyper/hypotonia, and abnormal neuroimaging [1, 5]. BLBS was initially categorized as a progressive neurodegenerative syndrome caused by de novo heterozygous variants in either H3-3A or H3-3B [1-4]. Here, we analyze the data of the 58 previously published individuals along 38 unpublished, unrelated individuals. In this larger cohort of 96 people, we identify causative missense, synonymous, and stop-loss variants. We also expand upon the phenotypic characterization by elaborating on the neurodevelopmental component of BLBS. Notably, phenotypic heterogeneity was present even amongst individuals harboring the same variant. To explore the complex phenotypic variation in this expanded cohort, the relationships between syndromic phenotypes with three variables of interest were interrogated: sex, gene containing the causative variant, and variant location in the H3.3 protein. While specific genotype-phenotype correlations have not been conclusively delineated, the results presented here suggest that the location of the variants within the H3.3 protein and the affected gene (H3-3A or H3-3B) contribute more to the severity of distinct phenotypes than sex. Since these variables do not account for all BLBS phenotypic variability, these findings suggest that additional factors may play a role in modifying the phenotypes of affected individuals. Histones are poised at the interface of genetics and epigenetics, highlighting the potential role for gene-environment interactions and the importance of future research.

Validation of 3 Computer-Aided Facial Phenotyping Tools (DeepGestalt, GestaltMatcher, and D-Score): Comparative Diagnostic Accuracy Study

BACKGROUND

While characteristic facial features provide important clues for finding the correct diagnosis in genetic syndromes, valid assessment can be challenging. The next-generation phenotyping algorithm DeepGestalt analyzes patient images and provides syndrome suggestions. GestaltMatcher matches patient images with similar facial features. The new D-Score provides a score for the degree of facial dysmorphism.

OBJECTIVE

We aimed to test state-of-the-art facial phenotyping tools by benchmarking GestaltMatcher and D-Score and comparing them to DeepGestalt.

METHODS

Using a retrospective sample of 4796 images of patients with 486 different genetic syndromes (London Medical Database, GestaltMatcher Database, and literature images) and 323 inconspicuous control images, we determined the clinical use of D-Score, GestaltMatcher, and DeepGestalt, evaluating sensitivity; specificity; accuracy; the number of supported diagnoses; and potential biases such as age, sex, and ethnicity.

RESULTS

DeepGestalt suggested 340 distinct syndromes and GestaltMatcher suggested 1128 syndromes. The top-30 sensitivity was higher for DeepGestalt (88%, SD 18%) than for GestaltMatcher (76%, SD 26%). DeepGestalt generally assigned lower scores but provided higher scores for patient images than for inconspicuous control images, thus allowing the 2 cohorts to be separated with an area under the receiver operating characteristic curve (AUROC) of 0.73. GestaltMatcher could not separate the 2 classes (AUROC 0.55). Trained for this purpose, D-Score achieved the highest discriminatory power (AUROC 0.86). D-Score's levels increased with the age of the depicted individuals. Male individuals yielded higher D-scores than female individuals. Ethnicity did not appear to influence D-scores.

CONCLUSIONS

If used with caution, algorithms such as D-score could help clinicians with constrained resources or limited experience in syndromology to decide whether a patient needs further genetic evaluation. Algorithms such as DeepGestalt could support diagnosing rather common genetic syndromes with facial abnormalities, whereas algorithms such as GestaltMatcher could suggest rare diagnoses that are unknown to the clinician in patients with a characteristic, dysmorphic face.

da Silva C, de Castro Michelassi G, Brentani H, Nunes FL, Machado-Lima A. Computer aided diagnosis of neurodevelopmental disorders and genetic syndromes based on facial images - A systematic literature review

Neurodevelopment disorders can result in facial dysmorphisms. Therefore, the analysis of facial images using image processing and machine learning techniques can help construct systems for diagnosing genetic syndromes and neurodevelopmental disorders. The systems offer faster and cost-effective alternatives for genotyping tests, particularly when dealing with large-scale applications. However, there are still challenges to overcome to ensure the accuracy and reliability of computer-aided diagnosis systems. This article presents a systematic review of such initiatives, including 55 articles. The main aspects used to develop these diagnostic systems were discussed, namely datasets - availability, type of image, size, ethnicities and syndromes - types of facial features, techniques used for normalization, dimensionality reduction and classification, deep learning, as well as a discussion related to the main gaps, challenges and opportunities.

Application of facial analysis Technology in Clinical Genetics: Considerations for diverse populations

Facial analysis technology in rare diseases has the potential to shorten the diagnostic odyssey by providing physicians with a valuable diagnostic tool. Given that most clinical genetic resources focus on populations of European descent, we compare craniofacial features in genetic syndromes across different populations and review how machine learning algorithms perform on diagnosing genetic syndromes in geographically and ethnically diverse populations. We also discuss the value of populations from ancestrally diverse backgrounds in the training set of machine learning algorithms. Finally, this review demonstrates that across diverse population groups, machine learning models have outstanding accuracy as supported by the area under the curve values greater than 0.9. Artificial intelligence is only in its infancy in the diagnosis of rare disease in diverse populations and will become more accurate as larger and more diverse training sets, including a wider spectrum of ages, particularly infants, are studied.

Usefulness of automated image analysis for recognition of the fragile X syndrome gestalt in Congolese subjects

BACKGROUND

Computer-aided software such as the facial image diagnostic aid (FIDA) and Face2Gene has been developed to perform pattern recognition of facial features with promising clinical results. The aim of this pilot study was to test Face2Gene's recognition performance on Bantu Congolese subjects with Fragile X syndrome (FXS) as compared to Congolese subjects with intellectual disability but without FXS (non-FXS).

METHOD

Frontal facial photograph from 156 participants (14 patients with FXS and 142 controls) predominantly young-adults to adults, median age 18.9 age range 4-39yo, were uploaded. Automated face analysis was conducted by using the technology used in proprietary software tools called Face2Gene CLINIC and Face2Gene RESEARCH (version 17.6.2). To estimate the statistical power of the Face2Gene technology in distinguishing affected individuals from controls, a cross validation scheme was used.

RESULTS

The similarity seen in the upper facial region (of males and females) is greater than the similarity seen in other parts of the face. Binary comparison of subjects with FXS versus non-FXS and subjects with FXS versus subjects with Down syndrome reveal an area under the curve values of 0.955 (p = 0.002) and 0.986 (p = 0.003).

CONCLUSION

The Face2Gene algorithm is separating well between FXS and Non-FXS subjects.

A Deep-Learning-Based Method Can Detect Both Common and Rare Genetic Disorders in Fetal Ultrasound

A global survey indicates that genetic syndromes affect approximately 8% of the population, but most genetic diagnoses can only be performed after babies are born. Abnormal facial characteristics have been identified in various genetic diseases; however, current facial identification technologies cannot be applied to prenatal diagnosis. We developed Pgds-ResNet, a fully automated prenatal screening algorithm based on deep neural networks, to detect high-risk fetuses affected by a variety of genetic diseases. In screening for Trisomy 21, Trisomy 18, Trisomy 13, and rare genetic diseases, Pgds-ResNet achieved sensitivities of 0.83, 0.92, 0.75, and 0.96, and specificities of 0.94, 0.93, 0.95, and 0.92, respectively. As shown in heatmaps, the abnormalities detected by Pgds-ResNet are consistent with clinical reports. In a comparative experiment, the performance of Pgds-ResNet is comparable to that of experienced sonographers. This fetal genetic screening technology offers an opportunity for early risk assessment and presents a non-invasive, affordable, and complementary method to identify high-risk fetuses affected by genetic diseases. Additionally, it has the capability to screen for certain rare genetic conditions, thereby enhancing the clinic's detection rate.

Loss-of-function variants in CUL3 cause a syndromic neurodevelopmental disorder

Purpose

De novo variants in CUL3 (Cullin-3 ubiquitin ligase) have been strongly associated with neurodevelopmental disorders (NDDs), but no large case series have been reported so far. Here we aimed to collect sporadic cases carrying rare variants in CUL3, describe the genotype-phenotype correlation, and investigate the underlying pathogenic mechanism.

Methods

Genetic data and detailed clinical records were collected via multi-center collaboration. Dysmorphic facial features were analyzed using GestaltMatcher. Variant effects on CUL3 protein stability were assessed using patient-derived T-cells.

Results

We assembled a cohort of 35 individuals with heterozygous CUL3 variants presenting a syndromic NDD characterized by intellectual disability with or without autistic features. Of these, 33 have loss-of-function (LoF) and two have missense variants. CUL3 LoF variants in patients may affect protein stability leading to perturbations in protein homeostasis, as evidenced by decreased ubiquitin-protein conjugates in vitro . Specifically, we show that cyclin E1 (CCNE1) and 4E-BP1 (EIF4EBP1), two prominent substrates of CUL3, fail to be targeted for proteasomal degradation in patient-derived cells.

Conclusion

Our study further refines the clinical and mutational spectrum of CUL3 -associated NDDs, expands the spectrum of cullin RING E3 ligase-associated neuropsychiatric disorders, and suggests haploinsufficiency via LoF variants is the predominant pathogenic mechanism.

Identifying the genetic causes of developmental disorders and intellectual disability in Africa: a systematic literature review

Objective: Genetic variants cause a significant portion of developmental disorders and intellectual disabilities (DD/ID), but clinical and genetic heterogeneity makes identification challenging. Compounding the issue is a lack of ethnic diversity in studies into the genetic aetiology of DD/ID, with a dearth of data from Africa. This systematic review aimed to comprehensively describe the current knowledge from the African continent on this topic. Method: Applicable literature published up until July 2021 was retrieved from PubMed, Scopus and Web of Science databases, following PRISMA guidelines, focusing on original research reports on DD/ID where African patients were the focus of the study. The quality of the dataset was assessed using appraisal tools from the Joanna Briggs Institute, whereafter metadata was extracted for analysis. Results: A total of 3,803 publications were extracted and screened. After duplicate removal, title, abstract and full paper screening, 287 publications were deemed appropriate for inclusion. Of the papers analysed, a large disparity was seen between work emanating from North Africa compared to sub-Saharan Africa, with North Africa dominating the publications. Representation of African scientists on publications was poorly balanced, with most research being led by international researchers. There are very few systematic cohort studies, particularly using newer technologies, such as chromosomal microarray and next-generation sequencing. Most of the reports on new technology data were generated outside Africa. Conclusion: This review highlights how the molecular epidemiology of DD/ID in Africa is hampered by significant knowledge gaps. Efforts are needed to produce systematically obtained high quality data that can be used to inform appropriate strategies to implement genomic medicine for DD/ID on the African continent, and to successfully bridge healthcare inequalities.

An automatic facial landmarking for children with rare diseases

Two to three thousand syndromes modify facial features: their screening requires the eye of an expert in dysmorphology. A widely used tool in shape characterization is geometric morphometrics based on landmarks, which are precise and reproducible anatomical points. Landmark positioning is user dependent and time consuming. Many automatic landmarking tools are currently available but do not work for children, because they have mainly been trained using photographic databases of healthy adults. Here, we developed a method for building an automatic landmarking pipeline for frontal and lateral facial photographs as well as photographs of external ears. We evaluated the algorithm on patients diagnosed with Treacher Collins (TC) syndrome as it is the most frequent mandibulofacial dysostosis in humans and is clinically recognizable although highly variable in severity. We extracted photographs from the photographic database of the maxillofacial surgery and plastic surgery department of Hôpital Necker-Enfants Malades in Paris, France with the diagnosis of TC syndrome. The control group was built from children admitted for craniofacial trauma or skin lesions. After testing two methods of object detection by bounding boxes, a Haar Cascade-based tool and a Faster Region-based Convolutional Neural Network (Faster R-CNN)-based tool, we evaluated three different automatic annotation algorithms: the patch-based active appearance model (AAM), the holistic AAM, and the constrained local model (CLM). The final error corresponding to the distance between the points placed by automatic annotation and those placed by manual annotation was reported. We included, respectively, 1664, 2044, and 1375 manually annotated frontal, profile, and ear photographs. Object recognition was optimized with the Faster R-CNN-based detector. The best annotation model was the patch-based AAM (p < 0.001 for frontal faces, p = 0.082 for profile faces and p < 0.001 for ears). This automatic annotation model resulted in the same classification performance as manually annotated data. Pretraining on public photographs did not improve the performance of the model. We defined a pipeline to create automatic annotation models adapted to faces with congenital anomalies, an essential prerequisite for research in dysmorphology.

Facial features of lysosomal storage disorders

INTRODUCTION

The use of facial recognition technology has diversified the diagnostic toolbelt for clinicians and researchers for the accurate diagnoses of patients with rare and challenging disorders. Specific identifiers in patient images can be grouped using artificial intelligence to allow the recognition of diseases and syndromes with similar features. Lysosomal storage disorders are rare, and some have prominent and unique features that may be used to train the accuracy of facial recognition software algorithms. Noteworthy features of lysosomal storage disorders (LSDs) include facial features such as prominent brows, wide noses, thickened lips, mouth, and chin, resulting in coarse and rounded facial features.

AREAS COVERED

We evaluated and report the prevalence of facial phenotypes in patients with different LSDs, noting two current examples when artificial intelligence strategies have been utilized to identify distinctive facies.

EXPERT OPINION

Specific LSDs, including Gaucher disease, Mucolipidosis IV and Fabry disease have recently been distinguished using facial recognition software. Additional lysosomal disorders LSDs lysosomal storage disorders with unique and distinguishable facial features also merit evaluation using this technology. These tools may ultimately aid in the identification of specific LSDs and shorten the diagnostic odyssey for patients with these rare and under-recognized disorders.

The genetic determinants of oral diseases in Africa: The gaps should be filled

Oral diseases are a major health concern and are among the most prevalent diseases globally. This problem is becoming more prominent in the rapidly growing populations of Africa. It is well documented that Africa exhibits the most diverse genetic make-up in the world. However, little work has been conducted to understand the genetic basis of oral diseases in Africans. Oral health is often neglected and receives low prioritisation from funders and governments. The genetic determinants of highly prevalent oral diseases such as dental caries and periodontal disease, and regionally prevalent conditions such as oral cancer and NOMA, are largely under-researched areas despite numerous articles alluding to a high burden of these diseases in African populations. Therefore, this review aims to shed light on the significant gaps in research on the genetic and genomic aspects of oral diseases in African populations and highlights the urgent need for evidence-based dentistry, in tandem with the development of the dentist/scientist workforce.

Algorithmic fairness in computational medicine

Machine learning models are increasingly adopted for facilitating clinical decision-making. However, recent research has shown that machine learning techniques may result in potential biases when making decisions for people in different subgroups, which can lead to detrimental effects on the health and well-being of specific demographic groups such as vulnerable ethnic minorities. This problem, termed algorithmic bias, has been extensively studied in theoretical machine learning recently. However, the impact of algorithmic bias on medicine and methods to mitigate this bias remain topics of active discussion. This paper presents a comprehensive review of algorithmic fairness in the context of computational medicine, which aims at improving medicine with computational approaches. Specifically, we overview the different types of algorithmic bias, fairness quantification metrics, and bias mitigation methods, and summarize popular software libraries and tools for bias evaluation and mitigation, with the goal of providing reference and insights to researchers and practitioners in computational medicine.

Vici syndrome in Israel: Clinical and molecular insights

Introduction: Vici Syndrome is a rare, severe, neurodevelopmental/neurodegenerative disorder with multi-systemic manifestations presenting in infancy. It is mainly characterized by global developmental delay, seizures, agenesis of the corpus callosum, hair and skin hypopigmentation, bilateral cataract, and varying degrees of immunodeficiency, among other features. Vici Syndrome is caused by biallelic pathogenic variants in EPG5, resulting in impaired autophagy. Thus far, the condition has been reported in less than a hundred individuals.

Objective and Methods: We aimed to characterize the clinical and molecular findings in individuals harboring biallelic EPG5 variants, recruited from four medical centers in Israel. Furthermore, we aimed to utilize a machine learning-based tool to assess facial features of Vici syndrome.

Results: Eleven cases of Vici Syndrome from five unrelated families, one of which was diagnosed prenatally with subsequent termination of pregnancy, were recruited. A total of five disease causing variants were detected in EPG5: two novel: c.2554-5A>G and c.1461delC; and 3 previously reported: c.3447G>A, c.5993C>G, and c.1007A>G, the latter previously identified in several patients of Ashkenazi-Jewish (AJ) descent. Amongst 140,491 individuals screened by the Dor Yeshorim Program, we show that the c.1007A>G variant has an overall carrier frequency of 0.45% (1 in 224) among AJ individuals. Finally, based on two-dimensional facial photographs of individuals with Vici syndrome (n = 19), a composite facial mask was created using the DeepGestalt algorithm, illustrating facial features typical of this disorder.

Conclusion: We report on ten children and one fetus from five unrelated families, affected with Vici syndrome, and describe prenatal and postnatal characteristics. Our findings contribute to the current knowledge regarding the molecular basis and phenotypic features of this rare syndrome. Additionally, the deep learning-based facial gestalt adds to the clinician’s diagnostic toolbox and may aid in facilitating identification of affected individuals.

Artificial Intelligence in Biological Sciences

Artificial intelligence (AI), currently a cutting-edge concept, has the potential to improve the quality of life of human beings. The fields of AI and biological research are becoming more intertwined, and methods for extracting and applying the information stored in live organisms are constantly being refined. As the field of AI matures with more trained algorithms, the potential of its application in epidemiology, the study of host–pathogen interactions and drug designing widens. AI is now being applied in several fields of drug discovery, customized medicine, gene editing, radiography, image processing and medication management. More precise diagnosis and cost-effective treatment will be possible in the near future due to the application of AI-based technologies. In the field of agriculture, farmers have reduced waste, increased output and decreased the amount of time it takes to bring their goods to market due to the application of advanced AI-based approaches. Moreover, with the use of AI through machine learning (ML) and deep-learning-based smart programs, one can modify the metabolic pathways of living systems to obtain the best possible outputs with the minimal inputs. Such efforts can improve the industrial strains of microbial species to maximize the yield in the bio-based industrial setup. This article summarizes the potentials of AI and their application to several fields of biology, such as medicine, agriculture, and bio-based industry.

Case Report: The success of face analysis technology in extremely rare genetic diseases in Korea: Tatton–Brown–Rahman syndrome and Say–Barber –Biesecker–Young–Simpson variant of ohdo syndrome

Tatton–Brown–Rahman syndrome (TBRS) and Say–Barber–Biesecker– Young–Simpson variant of Ohdo syndrome (SBBYSS) are extremely rare genetic disorders with less than 100 reported cases. Patients with these disorders exhibit a characteristic facial dysmorphism: TBRS is characterized by a round face, a straight and thick eyebrow, and prominent maxillary incisors, whereas SBBYSS is characterized by mask-like facies, blepharophimosis, and ptosis. The usefulness of Face2Gene as a tool for the identification of dysmorphology syndromes is discussed, because, in these patients, it suggested TBRS and SBBYSS within the top five candidate disorders. Face2Gene is useful for the diagnosis of extremely rare diseases in Korean patients, suggesting the possibility of expanding its clinical applications.

Facial imaging to screen for fetal alcohol spectrum disorder: A scoping review

Facial imaging tools have rapidly advanced in recent years and show potential for use in fetal alcohol spectrum disorder (FASD) screening and diagnosis. This scoping review describes the current state of evidence regarding the use of facial imaging being as a screening tool for FASD at a community level. This review follows the guidelines for the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) extension for scoping reviews and is registered with the Open Science Framework (osf.io/e4xw6). An electronic search of five databases was conducted. The time frame was limited to the period 2006 to 2022. The search included any form of imaging of the head, neck, oral cavity, and dentition. Animal and antenatal studies were excluded, as were those using only brain imaging. The search retrieved 730 unique titles. After title, abstract, and full-text screening, 28 primary studies were included in this review. Most studies were conducted with South African participants. Imaging included 2D photographs, 3D stereophotogrammetry, 3D laser scanning, and radiographs. Various measurements and landmarks were used to discriminate FASD from non-FASD participants, which included anthropometry, face shape analysis, and facial curvatures. Methods of data processing, analysis, and modeling ranged from manual methods to fully automated systems utilizing artificial intelligence. The use of facial imaging to screen for and diagnose patients with FASD is a rapidly advancing field. Most studies in the field remain exploratory, attempting to find accurate, reliable, and consistent landmarks and measures across different populations. For community screening, none of the tools in this review in their current form completely fulfill all the identified properties of an ideal screening tool. More research and development are needed prior to advocating for the use of any tool listed and the ethical implications are yet to be fully explored.

GestaltMatcher facilitates rare disease matching using facial phenotype descriptors

Many monogenic disorders cause a characteristic facial morphology. Artificial intelligence can support physicians in recognizing these patterns by associating facial phenotypes with the underlying syndrome through training on thousands of patient photographs. However, this 'supervised' approach means that diagnoses are only possible if the disorder was part of the training set. To improve recognition of ultra-rare disorders, we developed GestaltMatcher, an encoder for portraits that is based on a deep convolutional neural network. Photographs of 17,560 patients with 1,115 rare disorders were used to define a Clinical Face Phenotype Space, in which distances between cases define syndromic similarity. Here we show that patients can be matched to others with the same molecular diagnosis even when the disorder was not included in the training set. Together with mutation data, GestaltMatcher could not only accelerate the clinical diagnosis of patients with ultra-rare disorders and facial dysmorphism but also enable the delineation of new phenotypes.

A survey of U.S. public perspectives on facial recognition technology and facial imaging data practices in health and research contexts

Facial imaging and facial recognition technologies, now common in our daily lives, also are increasingly incorporated into health care processes, enabling touch-free appointment check-in, matching patients accurately, and assisting with the diagnosis of certain medical conditions. The use, sharing, and storage of facial data is expected to expand in coming years, yet little is documented about the perspectives of patients and participants regarding these uses. We developed a pair of surveys to gather public perspectives on uses of facial images and facial recognition technologies in healthcare and in health-related research in the United States. We used Qualtrics Panels to collect responses from general public respondents using two complementary and overlapping survey instruments; one focused on six types of biometrics (including facial images and DNA) and their uses in a wide range of societal contexts (including healthcare and research) and the other focused on facial imaging, facial recognition technology, and related data practices in health and research contexts specifically. We collected responses from a diverse group of 4,048 adults in the United States (2,038 and 2,010, from each survey respectively). A majority of respondents (55.5%) indicated they were equally worried about the privacy of medical records, DNA, and facial images collected for precision health research. A vignette was used to gauge willingness to participate in a hypothetical precision health study, with respondents split as willing to (39.6%), unwilling to (30.1%), and unsure about (30.3%) participating. Nearly one-quarter of respondents (24.8%) reported they would prefer to opt out of the DNA component of a study, and 22.0% reported they would prefer to opt out of both the DNA and facial imaging component of the study. Few indicated willingness to pay a fee to opt-out of the collection of their research data. Finally, respondents were offered options for ideal governance design of their data, as "open science"; "gated science"; and "closed science." No option elicited a majority response. Our findings indicate that while a majority of research participants might be comfortable with facial images and facial recognition technologies in healthcare and health-related research, a significant fraction expressed concern for the privacy of their own face-based data, similar to the privacy concerns of DNA data and medical records. A nuanced approach to uses of face-based data in healthcare and health-related research is needed, taking into consideration storage protection plans and the contexts of use.

Compound Heterozygote of Point Mutation and Chromosomal Microdeletion Involving OTUD6B Coinciding with ZMIZ1 Variant in Syndromic Intellectual Disability

The OTUD6B and ZMIZ1 genes were recently identified as causes of syndromic intellectual disability (ID) with shared phenotypes of facial dysmorphism, distal limb anomalies, and seizure disorders. OTUD6B- and ZMIZ1-related ID are inherited in autosomal recessive and autosomal dominant patterns, respectively. We report a 5-year-old girl with developmental delay, facial phenotypes resembling Williams syndrome, and cardiac defects. The patient also had terminal broadening of the fingers and polydactyly. Cytogenomic microarray (CMA), whole exome sequencing (WES), and mRNA analysis were performed. The CMA showed a paternally inherited 0.118 Mb deletion of 8q21.3, chr8:92084087–92202189, with OTUD6B involved. The WES identified a hemizygous OTUD6B variant, c.873delA (p.Lys291AsnfsTer3). The mother was heterozygous for this allele. The WES also demonstrated a heterozygous ZMIZ1 variant, c.1491 + 2T > C, in the patient and her father. This ZMIZ1 variant yielded exon 14 skipping, as evidenced by mRNA study. We suggest that Williams syndrome-like phenotypes, namely, periorbital edema, hanging cheek, and long and smooth philtrum represent expanded phenotypes of OTUD6B-related ID. Our data expand the genotypic spectrum of OTUD6B- and ZMIZ1-related disorders. This is the first reported case of a compound heterozygote featuring point mutation, chromosomal microdeletion of OTUD6B, and the unique event of OTUD6B, coupled with ZMIZ1 variants.

Deep Learning-Based Analysis of Face Images as a Screening Tool for Genetic Syndromes

Approximately 4% of the world’s population suffers from rare diseases. A vast majority of these disorders have a genetic background. The number of genes that have been linked to human diseases is constantly growing, but there are still genetic syndromes that remain to be discovered. The diagnostic yield of genetic testing is continuously developing, and the need for testing is becoming more significant. Due to limited resources, including trained clinical geneticists, patients referred to clinical genetics units must be accurately selected. Around 30–40% of genetic disorders are associated with specific facial characteristics called dysmorphic features. As part of our research, we analyzed the performance of classifiers based on deep learning face recognition models in detecting dysmorphic features. We tested two classification problems: a multiclass problem (15 genetic disorders vs. controls) and a two-class problem (disease vs. controls). In the multiclass task, the best result reached an accuracy level of 84%. The best accuracy result in the two-class problem reached 96%. More importantly, the binary classifier detected disease features in patients with diseases that were not previously present in the training dataset. The classifier was able to generalize differences between patients and controls, and to detect abnormalities without information about the specific disorder. This indicates that a screening tool based on deep learning and facial recognition could not only detect known diseases, but also detect patients with diseases that were not previously known. In the future, this tool could help in screening patients before they are referred to the genetic unit.

Facial analysis technology for the detection of Down syndrome in the Democratic Republic of the Congo

Down syndrome is one of the most common chromosomal anomalies affecting the world's population, with an estimated frequency of 1 in 700 live births. Despite its relatively high prevalence, diagnostic rates based on clinical features have remained under 70% for most of the developed world and even lower in countries with limited resources. While genetic and cytogenetic confirmation greatly increases the diagnostic rate, such resources are often non-existent in many low- and middle-income countries, particularly in Sub-Saharan Africa. To address the needs of countries with limited resources, the implementation of mobile, user-friendly and affordable technologies that aid in diagnosis would greatly increase the odds of success for a child born with a genetic condition. Given that the Democratic Republic of the Congo is estimated to have one of the highest rates of birth defects in the world, our team sought to determine if smartphone-based facial analysis technology could accurately detect Down syndrome in individuals of Congolese descent. Prior to technology training, we confirmed the presence of trisomy 21 using low-cost genomic applications that do not need advanced expertise to utilize and are available in many low-resourced countries. Our software technology trained on 132 Congolese subjects had a significantly improved performance (91.67% accuracy, 95.45% sensitivity, 87.88% specificity) when compared to previous technology trained on individuals who are not of Congolese origin (p < 5%). In addition, we provide the list of most discriminative facial features of Down syndrome and their ranges in the Congolese population. Collectively, our technology provides low-cost and accurate diagnosis of Down syndrome in the local population.

Facial recognition accuracy in photographs of Thai neonates with Down syndrome among physicians and the Face2Gene application

Down syndrome (DS) is typically recognizable in those who present with multiple dysmorphism, especially in regard to facial phenotypes. However, as the presentation of DS in neonates is less obvious, a phenotype-based presumptive diagnosis is more challenging. Recently, an artificial intelligence (AI) application, Face2Gene, was developed to help physicians recognize specific genetic syndromes by using two-dimensional facial photos. As of yet, there has not been any study comparing accuracy among physicians or applications. Our objective was to compare the facial recognition accuracy of DS in Thai neonates, using facial photographs, among physicians and the Face2Gene. Sixty-four Thai neonates at Thammasat University Hospital, with genetic testing and signed parental consent, were divided into a DS group (25) and non-DS group (39). Non-DS was further divided into unaffected (19) and those affected with other syndromes (20). Our results revealed physician accuracy (89%) was higher than the Face2Gene (81%); however, the application was higher in sensitivity (100%) than physicians (86%). While this application can serve as a helpful assistant in facilitating any genetic syndrome such as DS, to aid clinicians in recognizing DS facial features in neonates, it is not a replacement for well-trained doctors.

Williams syndrome

Williams syndrome (WS) is a relatively rare microdeletion disorder that occurs in as many as 1:7,500 individuals. WS arises due to the mispairing of low-copy DNA repetitive elements at meiosis. The deletion size is similar across most individuals with WS and leads to the loss of one copy of 25-27 genes on chromosome 7q11.23. The resulting unique disorder affects multiple systems, with cardinal features including but not limited to cardiovascular disease (characteristically stenosis of the great arteries and most notably supravalvar aortic stenosis), a distinctive craniofacial appearance, and a specific cognitive and behavioural profile that includes intellectual disability and hypersociability. Genotype-phenotype evidence is strongest for ELN, the gene encoding elastin, which is responsible for the vascular and connective tissue features of WS, and for the transcription factor genes GTF2I and GTF2IRD1, which are known to affect intellectual ability, social functioning and anxiety. Mounting evidence also ascribes phenotypic consequences to the deletion of BAZ1B, LIMK1, STX1A and MLXIPL, but more work is needed to understand the mechanism by which these deletions contribute to clinical outcomes. The age of diagnosis has fallen in regions of the world where technological advances, such as chromosomal microarray, enable clinicians to make the diagnosis of WS without formally suspecting it, allowing earlier intervention by medical and developmental specialists. Phenotypic variability is considerable for all cardinal features of WS but the specific sources of this variability remain unknown. Further investigation to identify the factors responsible for these differences may lead to mechanism-based rather than symptom-based therapies and should therefore be a high research priority.

The point-of-care use of a facial phenotyping tool in the genetics clinic: Enhancing diagnosis and education with machine learning

Computer-assisted pattern recognition platforms, such as Face2Gene® (F2G), can facilitate the diagnosis of children with rare genetic syndromes by comparing a patient's features to known genetic diagnoses. Our work designed, implemented, and evaluated an innovative model of care in clinical genetics in a heterogeneous and multicultural patient population that utilized this facial phenotyping software at the point-of-care. We assessed the performance of F2G by comparing the suggested diagnoses to the patient's confirmed molecular diagnosis. Providers' overall experiences with the technology and trainees' educational experiences were assessed with questionnaires. We achieved an overall diagnostic yield of 57%. This increased to 82% when cases diagnosed with syndromes not recognized by F2G were removed. The mean rank of a confirmed diagnosis in the top 10 was 2.3 (CI 1.5-3.2) and the mean gestalt score 37.6%. The most commonly suggested diagnoses were Noonan syndrome, mucopolysaccharidosis, and 22q11.2 deletion syndrome. Our qualitative assessment revealed that clinicians and trainees saw value using the tool in practice. Overall, this work helped to implement an innovative patient care delivery model in clinical genetics that utilizes a facial phenotyping tool at the point-of-care. Our data suggest that F2G has utility in the genetics clinic as a clinical decision support tool in diverse populations, with a majority of patients having their eventual diagnosis listed in the top 10 suggested syndromes based on a photograph alone. It shows promise for further integration into clinical care and medical education, and we advocate for its continued use, adoption and refinement along with transparent and accountable industrial partnerships.

Efficiency of Computer-Aided Facial Phenotyping (DeepGestalt) in Individuals With and Without a Genetic Syndrome: Diagnostic Accuracy Study

Background

Collectively, an estimated 5% of the population have a genetic disease. Many of them feature characteristics that can be detected by facial phenotyping. Face2Gene CLINIC is an online app for facial phenotyping of patients with genetic syndromes. DeepGestalt, the neural network driving Face2Gene, automatically prioritizes syndrome suggestions based on ordinary patient photographs, potentially improving the diagnostic process. Hitherto, studies on DeepGestalt’s quality highlighted its sensitivity in syndromic patients. However, determining the accuracy of a diagnostic methodology also requires testing of negative controls.

Objective

The aim of this study was to evaluate DeepGestalt's accuracy with photos of individuals with and without a genetic syndrome. Moreover, we aimed to propose a machine learning–based framework for the automated differentiation of DeepGestalt’s output on such images.

Methods

Frontal facial images of individuals with a diagnosis of a genetic syndrome (established clinically or molecularly) from a convenience sample were reanalyzed. Each photo was matched by age, sex, and ethnicity to a picture featuring an individual without a genetic syndrome. Absence of a facial gestalt suggestive of a genetic syndrome was determined by physicians working in medical genetics. Photos were selected from online reports or were taken by us for the purpose of this study. Facial phenotype was analyzed by DeepGestalt version 19.1.7, accessed via Face2Gene CLINIC. Furthermore, we designed linear support vector machines (SVMs) using Python 3.7 to automatically differentiate between the 2 classes of photographs based on DeepGestalt's result lists.

Results

We included photos of 323 patients diagnosed with 17 different genetic syndromes and matched those with an equal number of facial images without a genetic syndrome, analyzing a total of 646 pictures. We confirm DeepGestalt’s high sensitivity (top 10 sensitivity: 295/323, 91%). DeepGestalt’s syndrome suggestions in individuals without a craniofacially dysmorphic syndrome followed a nonrandom distribution. A total of 17 syndromes appeared in the top 30 suggestions of more than 50% of nondysmorphic images. DeepGestalt’s top scores differed between the syndromic and control images (area under the receiver operating characteristic [AUROC] curve 0.72, 95% CI 0.68-0.76; P<.001). A linear SVM running on DeepGestalt’s result vectors showed stronger differences (AUROC 0.89, 95% CI 0.87-0.92; P<.001).

Conclusions

DeepGestalt fairly separates images of individuals with and without a genetic syndrome. This separation can be significantly improved by SVMs running on top of DeepGestalt, thus supporting the diagnostic process of patients with a genetic syndrome. Our findings facilitate the critical interpretation of DeepGestalt’s results and may help enhance it and similar computer-aided facial phenotyping tools.

Rubinstein-Taybi syndrome in diverse populations

Rubinstein-Taybi syndrome (RSTS) is an autosomal dominant disorder, caused by loss-of-function variants in CREBBP or EP300. Affected individuals present with distinctive craniofacial features, broad thumbs and/or halluces, and intellectual disability. RSTS phenotype has been well characterized in individuals of European descent but not in other populations. In this study, individuals from diverse populations with RSTS were assessed by clinical examination and facial analysis technology. Clinical data of 38 individuals from 14 different countries were analyzed. The median age was 7 years (age range: 7 months to 47 years), and 63% were females. The most common phenotypic features in all population groups included broad thumbs and/or halluces in 97%, convex nasal ridge in 94%, and arched eyebrows in 92%. Face images of 87 individuals with RSTS (age range: 2 months to 47 years) were collected for evaluation using facial analysis technology. We compared images from 82 individuals with RSTS against 82 age- and sex-matched controls and obtained an area under the receiver operating characteristic curve (AUC) of 0.99 (p < .001), demonstrating excellent discrimination efficacy. The discrimination was, however, poor in the African group (AUC: 0.79; p = .145). Individuals with EP300 variants were more effectively discriminated (AUC: 0.95) compared with those with CREBBP variants (AUC: 0.93). This study shows that clinical examination combined with facial analysis technology may enable earlier and improved diagnosis of RSTS in diverse populations.

Hyperphosphatasia with mental retardation syndrome type 4 in three unrelated South African patients

Hyperphosphatasia with mental retardation syndrome (HPMRS) is a rare autosomal recessive disorder caused by pathogenic variants in genes involved in glycosylphosphatidylinositol metabolism that result in a similar phenotype. We describe the first three patients with HPMRS from sub-Saharan Africa. Detection was assisted by Face2Gene phenotype matching and confirmed by the presence of elevated serum alkaline phosphatase. All three patients had severe intellectual disability, absent speech, hypotonia and palatal abnormality (cleft palate in two, very high-arched palate in one), no or minimal brachytelephalangy, and high serum alkaline phosphatase levels. Additional findings included seizures in two, and brain imaging abnormalities in two. In all three patients HPMRS was a top-20 gestalt match using Face2Gene. The overall phenotype is consistent with descriptions in the literature of HPMRS type 4, although not specific to it. Whole exome sequencing in the index patient and his mother detected a candidate variant in a homozygous state in the index patient (PGAP3:c.557G>C, p.Arg186Thr) and heterozygous in the mother. Further variant interpretation indicated pathogenicity. Sanger sequencing of another two patients identified the same homozygous, pathogenic variant, confirming a diagnosis of HPMRS type 4. The shared homozygous variant in apparently unrelated families, and in the absence of consanguinity, suggests the possibility of genetic drift due to a population bottleneck effect, and further research is recommended.

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

Physical examinations are recommended as part of a comprehensive evaluation for individuals with neurodevelopmental disorders (NDDs), such as autism spectrum disorder (ASD) and attention-deficit/hyperactivity disorder. These examinations should include assessment for morphological variants. Previous studies have shown an increase in morphological variants in individuals with NDDs, particularly ASD, and that these variants may be present in greater amounts in individuals with genetic alterations. Unfortunately, assessment for morphological variants can be subjective and time-consuming, and require a high degree of clinical expertise. Therefore, objective, automated methods of morphological assessment are desirable. This study compared the use of Face2Gene, an automated tool to explore facial morphological variants, to clinical consensus assessment, using a cohort of N = 290 twins enriched for NDDs (n = 135 with NDD diagnoses). Agreement between automated and clinical assessments were satisfactory to complete (78.3-100%). In our twin sample, individuals with NDDs did not have greater numbers of facial morphological variants when compared to those with typical development, nor when controlling for shared genetic and environmental factors within twin pairs. Common facial morphological variants in those with and without NDDs were similar and included thick upper lip vermilion, abnormality of the nasal tip, long face, and upslanted palpebral fissure. We conclude that although facial morphological variants can be assessed reliably in NDDs with automated tools like Face2Gene, clinical utility is limited when just exploring the facial region. Therefore, currently, automated assessments may best complement, rather than replace, in-person clinical assessments.

Diversity and dysmorphology

PURPOSE OF REVIEW

Dysmorphic features result from errors in morphogenesis frequently associated with genetic syndromes. Recognizing patterns of dysmorphic features is a critical step in the diagnosis and management of human congenital anomalies and genetic syndromes. This review presents recent developments in genetic syndromes and their related dysmorphology in diverse populations.

RECENT FINDINGS

Clinical findings in patients with genetic syndromes differ in their heterogeneity across different population groups. Some genetic syndromes have variable features in different ethnicities, in part due to specific background exam characteristics such as flat facial profiles or nasal differences; however, other genetic syndromes are similar across different ethnicities. Facial analysis technology is accurate in diagnosing genetic syndromes in populations around the world and is a powerful adjunct to conventional clinical examination. This accuracy also reinforces the concept that genetic syndromes can and should be diagnosed in any ethnicity.

SUMMARY

The increasing amount of data from studies on genetic syndromes in diverse populations is significantly improving our knowledge and approach to dysmorphic patients from various ethnic backgrounds. Optimal management of genetic syndromes requires early diagnosis, including in developing countries.

Artificial intelligence in clinical and genomic diagnostics

Artificial intelligence (AI) is the development of computer systems that are able to perform tasks that normally require human intelligence. Advances in AI software and hardware, especially deep learning algorithms and the graphics processing units (GPUs) that power their training, have led to a recent and rapidly increasing interest in medical AI applications. In clinical diagnostics, AI-based computer vision approaches are poised to revolutionize image-based diagnostics, while other AI subtypes have begun to show similar promise in various diagnostic modalities. In some areas, such as clinical genomics, a specific type of AI algorithm known as deep learning is used to process large and complex genomic datasets. In this review, we first summarize the main classes of problems that AI systems are well suited to solve and describe the clinical diagnostic tasks that benefit from these solutions. Next, we focus on emerging methods for specific tasks in clinical genomics, including variant calling, genome annotation and variant classification, and phenotype-to-genotype correspondence. Finally, we end with a discussion on the future potential of AI in individualized medicine applications, especially for risk prediction in common complex diseases, and the challenges, limitations, and biases that must be carefully addressed for the successful deployment of AI in medical applications, particularly those utilizing human genetics and genomics data.

Enabling Global Clinical Collaborations on Identifiable Patient Data: The Minerva Initiative

The clinical utility of computational phenotyping for both genetic and rare diseases is increasingly appreciated; however, its true potential is yet to be fully realized. Alongside the growing clinical and research availability of sequencing technologies, precise deep and scalable phenotyping is required to serve unmet need in genetic and rare diseases. To improve the lives of individuals affected with rare diseases through deep phenotyping, global big data interrogation is necessary to aid our understanding of disease biology, assist diagnosis, and develop targeted treatment strategies. This includes the application of cutting-edge machine learning methods to image data. As with most digital tools employed in health care, there are ethical and data governance challenges associated with using identifiable personal image data. There are also risks with failing to deliver on the patient benefits of these new technologies, the biggest of which is posed by data siloing. The Minerva Initiative has been designed to enable the public good of deep phenotyping while mitigating these ethical risks. Its open structure, enabling collaboration and data sharing between individuals, clinicians, researchers and private enterprise, is key for delivering precision public health.

Differentiation of MISSLA and Fanconi anaemia by computer-aided image analysis and presentation of two novel MISSLA siblings

Variants in DONSON were recently identified as the cause of microcephaly, short stature, and limb abnormalities syndrome (MISSLA). The clinical spectra of MISSLA and Fanconi anaemia (FA) strongly overlap. For that reason, some MISSLA patients have been clinically diagnosed with FA. Here, we present the clinical data of siblings with MISSLA featuring a novel DONSON variant and summarize the current literature on MISSLA. Additionally, we perform computer-aided image analysis using the DeepGestalt technology to test how distinct the facial features of MISSLA and FA patients are. We show that MISSLA has a specific facial gestalt. Notably, we find that also FA patients feature facial characteristics recognizable by computer-aided image analysis. We conclude that computer-assisted image analysis improves diagnostic precision in both MISSLA and FA.

Evaluation of Face2Gene using facial images of patients with congenital dysmorphic syndromes recruited in Japan

An increasing number of genetic syndromes present a challenge to clinical geneticists. A deep learning-based diagnosis assistance system, Face2Gene, utilizes the aggregation of "gestalt," comprising data summarizing features of patients' facial images, to suggest candidate syndromes. Because Face2Gene's results may be affected by ethnicity and age at which training facial images were taken, the system performance for patients in Japan is still unclear. Here, we present an evaluation of Face2Gene using the following two patient groups recruited in Japan: Group 1 consisting of 74 patients with 47 congenital dysmorphic syndromes, and Group 2 consisting of 34 patients with Down syndrome. In Group 1, facial recognition failed for 4 of 74 patients, while 13-21 of 70 patients had a diagnosis for which Face2Gene had not been trained. Omitting these 21 patients, for 85.7% (42/49) of the remainder, the correct syndrome was identified within the top 10 suggested list. In Group 2, for the youngest facial images taken for each of the 34 patients, Down syndrome was successfully identified as the highest-ranking condition using images taken from newborns to those aged 25 years. For the oldest facial images taken at ≥20 years in each of 17 applicable patients, Down syndrome was successfully identified as the highest- and second-highest-ranking condition in 82.2% (14/17) and 100% (17/17) of the patients using images taken from 20 to 40 years. These results suggest that Face2Gene in its current format is already useful in suggesting candidate syndromes to clinical geneticists, using patients with congenital dysmorphic syndromes in Japan.

The facial dysmorphology analysis technology in intellectual disability syndromes related to defects in the histones modifiers

Genetic syndromes are frequently associated with Intellectual Disability (ID), as well as craniofacial dysmorphisms. A group of ID syndromes with typical abnormal face related to chromatin remodeling defects, have been recognized, coining the term chromatinopathies. This is a molecular heterogeneous subset of congenital disorders caused by mutations of the various components of the Chromatin-Marking System (CMS), including modifiers of DNA and chromatin remodelers. We performed a phenotypic study on a sample of 120 individuals harboring variants in genes codifying for the histones enzymes, using the DeepGestalt technology. Three experiments (two multiclass comparison experiments and a frontal face-crop analysis) were conducted, analyzing respectively a total of 181 pediatric images in the first comparison experiment and 180 in the second, all individuals belonging predominantly to Caucasian population. The classification results were expressed in terms of the area under the curve (AUC) of the receiver-operating-characteristic curve (ROC). Significant values of AUC and low p-values were registered for all syndromes in the three experiments, in comparison with each other, with other ID syndromes characterized by recognizable craniofacial dysmorphisms and with unaffected controls. Final findings indicated that this group of diseases is characterized by distinctive dysmorphisms, which result pathognomonic. A correct interrogation and use of adequate informatics aids, could become a valid support for clinicians.

Next-generation phenotyping using computer vision algorithms in rare genomic neurodevelopmental disorders

Purpose

The interpretation of genetic variants after genome-wide analysis is complex in heterogeneous disorders such as intellectual disability (ID). We investigate whether algorithms can be used to detect if a facial gestalt is present for three novel ID syndromes and if these techniques can help interpret variants of uncertain significance.

Methods

Facial features were extracted from photos of ID patients harboring a pathogenic variant in three novel ID genes (PACS1, PPM1D, and PHIP) using algorithms that model human facial dysmorphism, and facial recognition. The resulting features were combined into a hybrid model to compare the three cohorts against a background ID population.

Results

We validated our model using images from 71 individuals with Koolen–de Vries syndrome, and then show that facial gestalts are present for individuals with a pathogenic variant in PACS1 (p = 8 × 10⁻⁴), PPM1D (p = 4.65 × 10⁻²), and PHIP (p = 6.3 × 10⁻³). Moreover, two individuals with a de novo missense variant of uncertain significance in PHIP have significant similarity to the expected facial phenotype of PHIP patients (p < 1.52 × 10⁻²).

Conclusion

Our results show that analysis of facial photos can be used to detect previously unknown facial gestalts for novel ID syndromes, which will facilitate both clinical and molecular diagnosis of rare and novel syndromes.

Facial recognition software in clinical dysmorphology

PURPOSE OF REVIEW

The current review aims to discuss the incorporation of facial recognition software into the clinical practice of dysmorphology and medical genetics.

RECENT FINDINGS

Facial recognition software has improved the process of generating a differential diagnosis for rare genetic syndromes, and recent publications demonstrate utility in both research and clinical applications. Software programs are freely available to verified medical providers and can be incorporated into routine clinic encounters.

SUMMARY

As facial recognition software capabilities improve, two-dimensional image capture with artificial intelligence interpretation may become a useful tool within many areas of medicine. Geneticists and researchers can use such software to enhance their differential diagnoses, to study similarities and differences between patient cohorts, and to improve the interpretation of genomic data. Pediatricians and subspecialists may use tools to identify patients who may benefit from a genetic evaluation, and educators can use these tools to interest students in the study of dysmorphoplogy and genetic syndromes.

Recognizable phenotypes in CDG

Pattern recognition, using a group of characteristic, or discriminating features, is a powerful tool in metabolic diagnostic. A classic example of this approach is used in biochemical analysis of urine organic acid analysis, where the reporting depends more on the correlation of pertinent positive and negative findings, rather than on the absolute values of specific markers. Similar uses of pattern recognition in the field of biochemical genetics include the interpretation of data obtained by metabolomics, like glycomics, where a recognizable pattern or the presence of a specific glycan sub-fraction can lead to the direct diagnosis of certain types of congenital disorders of glycosylation. Another indispensable tool is the use of clinical pattern recognition-or syndromology-relying on careful phenotyping. While genomics might uncover variants not essential in the final clinical expression of disease, and metabolomics could point to a mixture of primary but also secondary changes in biochemical pathways, phenomics describes the clinically relevant manifestations and the full expression of the disease. In the current review we apply phenomics to the field of congenital disorders of glycosylation, focusing on recognizable differentiating findings in glycosylation disorders, characteristic dysmorphic features and malformations in PMM2-CDG, and overlapping patterns among the currently known glycosylation disorders based on their pathophysiological basis.

Advances in computer-assisted syndrome recognition by the example of inborn errors of metabolism

Significant improvements in automated image analysis have been achieved in recent years and tools are now increasingly being used in computer-assisted syndromology. However, the ability to recognize a syndromic facial gestalt might depend on the syndrome and may also be confounded by severity of phenotype, size of available training sets, ethnicity, age, and sex. Therefore, benchmarking and comparing the performance of deep-learned classification processes is inherently difficult. For a systematic analysis of these influencing factors we chose the lysosomal storage diseases mucolipidosis as well as mucopolysaccharidosis type I and II that are known for their wide and overlapping phenotypic spectra. For a dysmorphic comparison we used Smith-Lemli-Opitz syndrome as another inborn error of metabolism and Nicolaides-Baraitser syndrome as another disorder that is also characterized by coarse facies. A classifier that was trained on these five cohorts, comprising 289 patients in total, achieved a mean accuracy of 62%. We also developed a simulation framework to analyze the effect of potential confounders, such as cohort size, age, sex, or ethnic background on the distinguishability of phenotypes. We found that the true positive rate increases for all analyzed disorders for growing cohorts (n = [10...40]) while ethnicity and sex have no significant influence. The dynamics of the accuracies strongly suggest that the maximum distinguishability is a phenotype-specific value, which has not been reached yet for any of the studied disorders. This should also be a motivation to further intensify data sharing efforts, as computer-assisted syndrome classification can still be improved by enlarging the available training sets.