A Radiation-Free Classification Pipeline for Craniosynostosis Using Statistical Shape Modeling

New diagnostic criteria for metopic ridges and trigonocephaly: a 3D geometric approach

BACKGROUND

Trigonocephaly occurs due to the premature fusion of the metopic suture, leading to a triangular forehead and hypotelorism. This condition often requires surgical correction for morphological and functional indications. Metopic ridges also originate from premature metopic closure but are only associated with mid-frontal bulging; their surgical correction is rarely required. Differential diagnosis between these two conditions can be challenging, especially in minor trigonocephaly.

METHODS

Two hundred seven scans of patients with trigonocephaly (90), metopic rigdes (27), and controls (90) were collected. Geometric morphometrics were used to quantify skull and orbital morphology as well as the interfrontal angle and the cephalic index. An innovative method was developed to automatically compute the frontal curvature along the metopic suture. Different machine-learning algorithms were tested to assess the predictive power of morphological data in terms of classification.

RESULTS

We showed that control patients, trigonocephaly and metopic rigdes have distinctive skull and orbital shapes. The 3D frontal curvature enabled a clear discrimination between groups (sensitivity and specificity > 92%). Furthermore, we reached an accuracy of 100% in group discrimination when combining 6 univariate measures.

CONCLUSION

Two diagnostic tools were proposed and demonstrated to be successful in assisting differential diagnosis for patients with trigonocephaly or metopic ridges. Further clinical assessments are required to validate the practical clinical relevance of these tools.

Quantifying dysmorphologies of the neurocranium using artificial neural networks

BACKGROUND

Craniosynostosis, a congenital condition characterized by the premature fusion of cranial sutures, necessitates objective methods for evaluating cranial morphology to enhance patient treatment. Current subjective assessments often lead to inconsistent outcomes. This study introduces a novel, quantitative approach to classify craniosynostosis and measure its severity.

METHODS

An artificial neural network was trained to classify normocephalic, trigonocephalic, and scaphocephalic head shapes based on a publicly available dataset of synthetic 3D head models. Each 3D model was converted into a low-dimensional shape representation based on the distribution of normal vectors, which served as the input for the neural network, ensuring complete patient anonymity and invariance to geometric size and orientation. Explainable AI methods were utilized to highlight significant features when making predictions. Additionally, the Feature Prominence (FP) score was introduced, a novel metric that captures the prominence of distinct shape characteristics associated with a given class. Its relationship with clinical severity scores was examined using the Spearman Rank Correlation Coefficient.

RESULTS

The final model achieved excellent test accuracy in classifying the different cranial shapes from their low-dimensional representation. Attention maps indicated that the network's attention was predominantly directed toward the parietal and temporal regions, as well as toward the region signifying vertex depression in scaphocephaly. In trigonocephaly, features around the temples were most pronounced. The FP score showed a strong positive monotonic relationship with clinical severity scores in both scaphocephalic (ρ = 0.83, p < 0.001) and trigonocephalic (ρ = 0.64, p < 0.001) models. Visual assessments further confirmed that as FP values rose, phenotypic severity became increasingly evident.

CONCLUSION

This study presents an innovative and accessible AI-based method for quantifying cranial shape that mitigates the need for adjustments due to age-specific size variations or differences in the spatial orientation of the 3D images, while ensuring complete patient privacy. The proposed FP score strongly correlates with clinical severity scores and has the potential to aid in clinical decision-making and facilitate multi-center collaborations. Future work will focus on validating the model with larger patient datasets and exploring the potential of the FP score for broader applications. The publicly available source code facilitates easy implementation, aiming to advance craniofacial care and research.

Machine learning applications in craniosynostosis diagnosis and treatment prediction: a systematic review

Craniosynostosis refers to the premature fusion of one or more of the fibrous cranial sutures connecting the bones of the skull. Machine learning (ML) is an emerging technology and its application to craniosynostosis detection and management is underexplored. This systematic review aims to evaluate the application of ML techniques in the diagnosis, severity assessment, and predictive modeling of craniosynostosis. A comprehensive search was conducted on the PubMed and Google Scholar databases using predefined keywords related to craniosynostosis and ML. Inclusion criteria encompassed peer-reviewed studies in English that investigated ML algorithms in craniosynostosis diagnosis, severity assessment, or treatment outcome prediction. Three independent reviewers screened the search results, performed full-text assessments, and extracted data from selected studies using a standardized form. Thirteen studies met the inclusion criteria and were included in the review. Of the thirteen papers examined on the application of ML to the identification and treatment of craniosynostosis, two papers were dedicated to sagittal craniosynostosis, five papers utilized several different types of craniosynostosis in the training and testing of their ML models, and six papers were dedicated to metopic craniosynostosis. ML models demonstrated high accuracy in identifying different types of craniosynostosis and objectively quantifying severity using innovative metrics such as metopic severity score and cranial morphology deviation. The findings highlight the significant strides made in utilizing ML techniques for craniosynostosis diagnosis, severity assessment, and predictive modeling. Predictive modeling of treatment outcomes following surgical interventions showed promising results, aiding in personalized treatment strategies. Despite methodological diversities among studies, the collective evidence underscores ML's transformative potential in revolutionizing craniosynostosis management.

Impact of data synthesis strategies for the classification of craniosynostosis

Introduction

Photogrammetric surface scans provide a radiation-free option to assess and classify craniosynostosis. Due to the low prevalence of craniosynostosis and high patient restrictions, clinical data are rare. Synthetic data could support or even replace clinical data for the classification of craniosynostosis, but this has never been studied systematically.

Methods

We tested the combinations of three different synthetic data sources: a statistical shape model (SSM), a generative adversarial network (GAN), and image-based principal component analysis for a convolutional neural network (CNN)-based classification of craniosynostosis. The CNN is trained only on synthetic data but is validated and tested on clinical data.

Results

The combination of an SSM and a GAN achieved an accuracy of 0.960 and an F1 score of 0.928 on the unseen test set. The difference to training on clinical data was smaller than 0.01. Including a second image modality improved classification performance for all data sources.

Conclusions

Without a single clinical training sample, a CNN was able to classify head deformities with similar accuracy as if it was trained on clinical data. Using multiple data sources was key for a good classification based on synthetic data alone. Synthetic data might play an important future role in the assessment of craniosynostosis.

3D-2D Distance Maps Conversion Enhances Classification of Craniosynostosis

OBJECTIVE

Diagnosis of craniosynostosis using photogrammetric 3D surface scans is a promising radiation-free alternative to traditional computed tomography. We propose a 3D surface scan to 2D distance map conversion enabling the usage of the first convolutional neural networks (CNNs)-based classification of craniosynostosis. Benefits of using 2D images include preserving patient anonymity, enabling data augmentation during training, and a strong under-sampling of the 3D surface with good classification performance.

METHODS

The proposed distance maps sample 2D images from 3D surface scans using a coordinate transformation, ray casting, and distance extraction. We introduce a CNN-based classification pipeline and compare our classifier to alternative approaches on a dataset of 496 patients. We investigate into low-resolution sampling, data augmentation, and attribution mapping.

RESULTS

Resnet18 outperformed alternative classifiers on our dataset with an F1-score of 0.964 and an accuracy of 98.4%. Data augmentation on 2D distance maps increased performance for all classifiers. Under-sampling allowed 256-fold computation reduction during ray casting while retaining an F1-score of 0.92. Attribution maps showed high amplitudes on the frontal head.

CONCLUSION

We demonstrated a versatile mapping approach to extract a 2D distance map from the 3D head geometry increasing classification performance, enabling data augmentation during training on 2D distance maps, and the usage of CNNs. We found that low-resolution images were sufficient for a good classification performance.

SIGNIFICANCE

Photogrammetric surface scans are a suitable craniosynostosis diagnosis tool for clinical practice. Domain transfer to computed tomography seems likely and can further contribute to reducing ionizing radiation exposure for infants.

Mesh2SSM: From Surface Meshes to Statistical Shape Models of Anatomy

Statistical shape modeling is the computational process of discovering significant shape parameters from segmented anatomies captured by medical images (such as MRI and CT scans), which can fully describe subject-specific anatomy in the context of a population. The presence of substantial non-linear variability in human anatomy often makes the traditional shape modeling process challenging. Deep learning techniques can learn complex non-linear representations of shapes and generate statistical shape models that are more faithful to the underlying population-level variability. However, existing deep learning models still have limitations and require established/optimized shape models for training. We propose Mesh2SSM, a new approach that leverages unsupervised, permutation-invariant representation learning to estimate how to deform a template point cloud to subject-specific meshes, forming a correspondence-based shape model. Mesh2SSM can also learn a population-specific template, reducing any bias due to template selection. The proposed method operates directly on meshes and is computationally efficient, making it an attractive alternative to traditional and deep learning-based SSM approaches.

Geometric learning and statistical modeling for surgical outcomes evaluation in craniosynostosis using 3D photogrammetry

BACKGROUND AND OBJECTIVE

Accurate and repeatable detection of craniofacial landmarks is crucial for automated quantitative evaluation of head development anomalies. Since traditional imaging modalities are discouraged in pediatric patients, 3D photogrammetry has emerged as a popular and safe imaging alternative to evaluate craniofacial anomalies. However, traditional image analysis methods are not designed to operate on unstructured image data representations such as 3D photogrammetry.

METHODS

We present a fully automated pipeline to identify craniofacial landmarks in real time, and we use it to assess the head shape of patients with craniosynostosis using 3D photogrammetry. To detect craniofacial landmarks, we propose a novel geometric convolutional neural network based on Chebyshev polynomials to exploit the point connectivity information in 3D photogrammetry and quantify multi-resolution spatial features. We propose a landmark-specific trainable scheme that aggregates the multi-resolution geometric and texture features quantified at every vertex of a 3D photogram. Then, we embed a new probabilistic distance regressor module that leverages the integrated features at every point to predict landmark locations without assuming correspondences with specific vertices in the original 3D photogram. Finally, we use the detected landmarks to segment the calvaria from the 3D photograms of children with craniosynostosis, and we derive a new statistical index of head shape anomaly to quantify head shape improvements after surgical treatment.

RESULTS

We achieved an average error of 2.74 ± 2.70 mm identifying Bookstein Type I craniofacial landmarks, which is a significant improvement compared to other state-of-the-art methods. Our experiments also demonstrated a high robustness to spatial resolution variability in the 3D photograms. Finally, our head shape anomaly index quantified a significant reduction of head shape anomalies as a consequence of surgical treatment.

CONCLUSION

Our fully automated framework provides real-time craniofacial landmark detection from 3D photogrammetry with state-of-the-art accuracy. In addition, our new head shape anomaly index can quantify significant head phenotype changes and can be used to quantitatively evaluate surgical treatment in patients with craniosynostosis.

Photogrammetry Applied to Neurosurgery: A Literature Review

Photogrammetry refers to the process of creating 3D models and taking measurements through the use of photographs. Photogrammetry has many applications in neurosurgery, such as creating 3D anatomical models and diagnosing and evaluating head shape and posture deformities. This review aims to summarize the uses of the technique in the neurosurgical practice and showcase the systems and software required for its implementation. A literature review was done in the online database PubMed. Papers were searched using the keywords "photogrammetry", "neurosurgery", "neuroanatomy", "craniosynostosis" and "scoliosis". The identified articles were later put through primary (abstracts and titles) and secondary (full text) screening for eligibility for inclusion. In total, 86 articles were included in the review from 315 papers identified. The review showed that the main uses of photogrammetry in the field of neurosurgery are related to the creation of 3D models of complex neuroanatomical structures and surgical approaches, accompanied by the uses for diagnosis and evaluation of patients with structural deformities of the head and trunk, such as craniosynostosis and scoliosis. Additionally, three instances of photogrammetry applied for more specific aims, namely, cervical spine surgery, skull-base surgery, and radiosurgery, were identified. Information was extracted on the software and systems used to execute the method. With the development of the photogrammetric method, it has become possible to create accurate 3D models of physical objects and analyze images with dedicated software. In the neurosurgical setting, this has translated into the creation of anatomical teaching models and surgical 3D models as well as the evaluation of head and spine deformities. Through those applications, the method has the potential to facilitate the education of residents and medical students and the diagnosis of patient pathologies.

Assessing craniofacial growth and form without landmarks: A new automatic approach based on spectral methods

We present a novel method for the morphometric analysis of series of 3D shapes, and demonstrate its relevance for the detection and quantification of two craniofacial anomalies: trigonocephaly and metopic ridges, using CT-scans of young children. Our approach is fully automatic, and does not rely on manual landmark placement and annotations. Our approach furthermore allows to differentiate shape classes, enabling successful differential diagnosis between trigonocephaly and metopic ridges, two related conditions characterized by triangular foreheads. These results were obtained using recent developments in automatic nonrigid 3D shape correspondence methods and specifically spectral approaches based on the functional map framework. Our method can capture local changes in geometric structure, in contrast to methods based, for instance, on global shape descriptors. As such, our approach allows to perform automatic shape classification and provides visual feedback on shape regions associated with different classes of deformations. The flexibility and generality of our approach paves the way for the application of spectral methods in quantitative medicine.