Uabundit Nongnut, Chaiyamoon Arada, Iamsaard Sitthichai, Yurasakpong Laphatrada, Nantasenamat Chanin, Suwannakhan Athikhun, Phunchago Nichapa
Department of Anatomy, Faculty of Medicine, Khon Kaen University, Khon Kaen 40002, Thailand.
In Silico and Clinical Anatomy Research Group (iSCAN), Department of Anatomy, Faculty of Science, Mahidol University, Bangkok 10400, Thailand.
Medicina (Kaunas). 2021 Nov 21;57(11):1282. doi: 10.3390/medicina57111282.
The landmark for neurosurgical approaches to access brain lesion is the pterion. The aim of the present study is to classify and examine the prevalence of all types of pterion variations and perform morphometric measurements from previously defined anthropological landmarks. One-hundred and twenty-four Thai dried skulls were investigated. Classification and morphometric measurement of the pterion was performed. Machine learning models were also used to interpret the morphometric findings with respect to sex and age estimation. Spheno-parietal type was the most common type (62.1%), followed by epipteric (11.7%), fronto-temporal (5.2%) and stellate (1.2%). Complete synostosis of the pterion suture was present in 18.5% and was only present in males. While most morphometric measurements were similar between males and females, the distances from the pterion center to the mastoid process and to the external occipital protuberance were longer in males. Random forest algorithm could predict sex with 80.7% accuracy (root mean square error = 0.38) when the pterion morphometric data were provided. Correlational analysis indicated that the distances from the pterion center to the anterior aspect of the frontozygomatic suture and to the zygomatic angle were positively correlated with age, which may serve as basis for age estimation in the future. Further studies are needed to explore the use of machine learning in anatomical studies and morphometry-based sex and age estimation. Thorough understanding of the anatomy of the pterion is clinically useful when planning pterional craniotomy, particularly when the position of the pterion may change with age.
用于脑部病变神经外科手术入路的标志性结构是翼点。本研究的目的是对所有类型的翼点变异进行分类并检查其发生率，并从先前定义的人类学标志进行形态测量。对124个泰国干燥颅骨进行了研究。对翼点进行了分类和形态测量。还使用机器学习模型来解释关于性别和年龄估计的形态测量结果。蝶顶型是最常见的类型（62.1%），其次是颞蝶型（11.7%）、额颞型（5.2%）和星型（1.2%）。翼点缝完全融合的情况在18.5%的个体中存在，且仅见于男性。虽然大多数形态测量在男性和女性之间相似，但男性从翼点中心到乳突和枕外隆凸的距离更长。当提供翼点形态测量数据时，随机森林算法预测性别的准确率为80.7%（均方根误差 = 0.38）。相关性分析表明，从翼点中心到颧额缝前缘和颧角的距离与年龄呈正相关，这可能为未来的年龄估计提供依据。需要进一步研究以探索机器学习在解剖学研究以及基于形态测量的性别和年龄估计中的应用。在计划翼点开颅手术时，尤其是当翼点的位置可能随年龄变化时，深入了解翼点的解剖结构在临床上很有用。
