A range of services are available to help with physical, developmental and behavioural issues. Our research shows that even lower levels of prenatal alcohol exposure can be linked to specific facial changes that persist into early childhood. Our aim was to study the potential risks of lower levels of prenatal alcohol exposure. This makes it difficult to predict the exact effects of drinking alcohol during pregnancy for any individual. While some deformities of FAS may be evident through prenatal ultrasound, it is difficult to diagnose FAS during pregnancy. The diagnosis is not based on a single symptom, and mild cases may be difficult to diagnose.

Signature Visualizations Reveal Individual Facial Dysmorphism

• The datasets examined in this work differ greatly in size and quality, which has a direct impact on their appropriateness for training AI models.
• During pregnancy, it is often used to monitor fetal growth, detect anomalies, determine the baby’s position, and examine the health of the placenta.
• Deep learning algorithms have showed promising results in a different types of medical applications, including the diagnosis of fetal facial defects.
• Figure 2C shows portrait and profile views of a child with FAS, selected because he has classic features as well as idiosyncratic mild proptosis.
• However, we did find weaker connections between different brain areas in children who were exposed to alcohol throughout pregnancy.
• These models have demonstrated high accuracy in identifying conditions such as cleft lip, palate, and micrognathia.

Lifelong treatment is required and is more effective if collaborative care coordination occurs between all professional agencies. The families of people with FAS should also be included in treatment interventions. Alcohol seems most damaging in the first trimester (three months) of pregnancy but can affect the fetus at any time during the pregnancy. Although more research is necessary, some studies show that the craniofacial differences of people with FAS may improve during or after adolescence. The traits most likely to persist are a thin upper lip and Substance abuse a smaller head circumference.

Links to NCBI Databases

• A range of services are available to help with physical, developmental and behavioural issues.
• Fetal magnetic resonance imaging (MRI) is used when more detailed information concerning face malformations is necessary.
• Autoencoders have been used to extract features and detect anomalies in fetal face pictures.
• A woman who drinks alcohol while she is pregnant may harm her developing baby (fetus).
• Because many people do not know they are pregnant during those first few weeks, the risk of FAS increases if you drink alcohol and have unprotected sex.

Reducing the quantity of labeled data necessary for training by reusing the learnt features from these models. Rather than fine-tuning the entire pre-trained model, another option is to extract deep features from the network’s intermediate layers. These attributes can be used to indicate facial characteristics at a high level. Obtain good performance even with less labeled data by training a separate classifier utilising these features on fetal anomalies dataset. Domain adaptation addresses the problem of transferring knowledge from the source domain (adult facial images) to a target domain (fetal facial images) when data distributions may differ (Chen et al. 2018).

Neurodevelopmental Symptoms

This combined data can help create a more accurate diagnosis as well as suggest suitable management and counselling alternatives. The detection of fetal facial defects using machine learning algorithms is a new field with prospective implications in prenatal diagnostics and healthcare (Barrag´an-Montero et al. 2021). Machine learning mainly works on different steps, The first stage in creating a machine learning model for detecting fetal face anomalies is to collect a large dataset.

This review explores the growing need for AI-based algorithms in diagnosing fetal facial anomalies, which are often difficult to detect due to limitations in current imaging techniques like ultrasound and MRI. These challenges include low resolution, motion artifacts, and insufficient annotated data, which hinder early and accurate diagnosis. AI algorithms, particularly deep learning models like Convolutional Neural Networks (CNNs) and U-Net, offers significant potential to overcome these challenges by analyzing large datasets and improving image analysis. Early diagnosis of these anomalies is crucial for enabling timely interventions, personalized treatment plans, and better prenatal care.

The AI system can provide visual representations of the discovered facial defects, emphasising the damaged regions or components. This enables healthcare professionals and patients to visualise the irregularities, which promotes in the development of faith in the system’s capabilities. Reviews that expressly address embryonic facial malformations are uncommon, but they give useful context for the current investigation.

• This trend indicates a consistent interest in improving traditional medical approaches to fetal anomaly detection.
• The comprehensive study intends to investigate the recent AI-based algorithms in this domain, highlighting their promise, limitations, and future perspectives.
• Aboughalia et al. (2021) examined the use of ultrasonography and MRI to detect embryonic facial anomalies, noting the complementing capabilities of both techniques.
• If a pregnant person has a problem with using alcohol and cannot stop using, substance use treatment during pregnancy should be offered.

These automated systems can speed up the process of detecting facial anomalies, saving time on analysis and interpretation. This is especially useful in crowded clinical environments, when healthcare personnel frequently have limited time for each patient. Furthermore, AI-based prenatal US has the potential to improve clinical quality control and medical resource imbalance, hence shortening the training cycle of young doctors (He et al. 2021). Deep learning approaches are more complicated, open source and publicly available.

Whole-Body Effects of Fetal Alcohol Syndrome

Research shows that alcohol exposure at specific times during pregnancy can affect the brain in various ways, resulting in a spectrum of brain disorders. Many individuals with prenatal alcohol exposure (PAE) who are seen in pediatric practices manifest developmental and behavioral challenges. Differential diagnoses for an individual with PAE can involve additional testing to determine growth deficits and dysmorphic facial features related to other syndromes, and to identify co-occurring conditions.

• People with FAS may also have a distinctive philtrum, which is the groove between the bottom of the nose and the top of the upper lip.
• Another limitation is an insufficient examination of the clinical consequences and real-world use of AI-based approaches.
• Because fetal facial anomalies are uncommon, it is challenging to collect required labeled data for algorithm development.

These approaches are intended to detect and assess potential facial defects in prenatal newborns. Traditional diagnostic methods for fetal facial defects include ultrasound imaging, fetal magnetic resonance imaging drunken fetal syndrome (MRI), amiocentesis, chorionic villus sampling (CVS), and maternal serum screening (Shafi et al. 2013). The frequently used Diagnostic procedures for fetal anomaly detections are Ultrasound and MRI, examples of ultrasound and MRI are shown in Fig.

Using AI to detect fetal facial anomalies is a difficult research that necessitates powerful algorithms and models. Explainable AI strategies can be used to increase confidence and acceptability in this domain. The ability of an AI system to provide clear explanations or justifications for its actions or predictions is referred to as explainable AI (Ullah et al. 2024; Oprescu et al. 2022). The algorithm can provide transparent reasons for its decisions by specifying specific rules for recognising fetal face anomalies.