Incorporation of a biparietal narrowing metric to improve the ability of machine learning models to detect sagittal craniosynostosis with 2D photographs

Abstract

OBJECTIVE Sagittal craniosynostosis is the most common form of craniosynostosis and typically results in scaphocephaly, which is characterized by biparietal narrowing, compensatory frontal bossing, and an occipital prominence. The cephalic index (CI) is a simple metric for quantifying the degree of cranial narrowing and is often used to diagnose sagittal craniosynostosis. However, patients with variant forms of sagittal craniosynostosis may present with a "normal" CI, depending on the part of the suture that is closed. As machine learning (ML) algorithms are developed to assist in the diagnosis of cranial deformities, metrics that reflect the other phenotypic features of sagittal craniosynostosis are needed. In this study the authors sought to describe the posterior arc angle (PAA), a measurement of biparietal narrowing that is obtained with 2D photographs, and elucidate the role of PAA as an adjuvant to the CI in characterizing scaphocephaly and the potential relevance of PAA in new ML model development. METHODS The authors retrospectively reviewed 1013 craniofacial patients treated during the period from 2006 to 2021. Orthogonal top-down photographs were used to calculate the CI and PAA. Distribution densities, receiver operating characteristic (ROC) curves, and chi-square analyses were used to describe the relative predictive utility of each method for sagittal craniosynostosis. RESULTS In total, 1001 patients underwent paired CI and PAA measurements and a clinical head shape diagnosis (sagittal craniosynostosis, n = 122; other cranial deformity, n = 565; normocephalic, n = 314). The area under the ROC curve (AUC) for the CI was 98.5% (95% confidence interval 97.8%–99.2%, p < 0.001), with an optimum specificity of 92.6% and sensitivity of 93.4%. The PAA had an AUC of 97.4% (95% confidence interval 96.0%–98.8%, p < 0.001) with an optimum specificity of 94.9% and sensitivity of 90.2%. In 6 of 122 (4.9%) cases of sagittal craniosynostosis, the PAA was abnormal while the CI was normal. This means that adding a PAA cutoff branch to a partition model increases the detection of sagittal craniosynostosis. CONCLUSIONS Both CI and PAA are excellent discriminators for sagittal craniosynostosis. Using an accuracy-optimized partition model, the addition of the PAA to the CI increased model sensitivity compared to using the CI alone. Using a model that incorporates both CI and PAA could assist in the early identification and treatment of sagittal craniosynostosis via automated and semiautomated algorithms that utilize tree-based ML models.

Artificial Intelligence Applications in Diagnosing and Managing Non-syndromic Craniosynostosis: A Comprehensive Review

“…The impact of accurate input data was clearly illustrated by You et al, who found models with input data from senior surgeons to be more accurate than those from junior surgeons [13]. Second, a large set of training data verified by multiple surgeons is needed to ensure reliability which is a very time-consuming endeavour [7][8][9][10][11][12][13][14]. Several different craniometrics to classify craniosynostosis have been proposed in these studies.…”

“…When using their ML model of support vector regression, they were able to identify brachycephaly and plagiocephaly with 86.7% accuracy [ 8 ]. Likewise, Anderson et al demonstrated the use of 2D photographs to measure the posterior arc angle and the cephalic index [ 9 ]. These have been found to accurately classify sagittal craniosynostosis and have the potential to be incorporated into ML models [ 9 ].…”

“…Likewise, Anderson et al demonstrated the use of 2D photographs to measure the posterior arc angle and the cephalic index [ 9 ]. These have been found to accurately classify sagittal craniosynostosis and have the potential to be incorporated into ML models [ 9 ]. Jong et al also showed similar results using a deep learning algorithm and 3D photogrammetry to sub-classify patients with 99.5% accuracy [ 10 ].…”

“…Several studies have used a similar methodology of collecting photographs or CT scans that are processed using programmes such as GrabCut and MATLAB to isolate the head shape from the images [7][8][9][10][11][12][13]. A test dataset, classified by expert physicians, is used to train the ML algorithm to measure cranial indices which are then used to classify the type of synostosis [7][8][9][10][11][12][13]. The accuracy of the algorithm is then assessed by running it on a separate dataset [7][8][9][10][11][12][13].…”

“…A test dataset, classified by expert physicians, is used to train the ML algorithm to measure cranial indices which are then used to classify the type of synostosis [7][8][9][10][11][12][13]. The accuracy of the algorithm is then assessed by running it on a separate dataset [7][8][9][10][11][12][13]. This is demonstrated in Table 1 and Figure 1.…”