Coutinho Catarina P, Zanchetta Ferdinando, Carbonelli Michele, Battista Marco, Galzignato Alice, La Morgia Chiara, Amore Giulia, Romagnoli Martina, Savini Giacomo, Brotto Luigi, Nucci Paolo, Caporali Leonardo, Bandello Francesco, Carelli Valerio, Cascavilla Maria Lucia, Fioresi Rita, Barboni Piero
Department of Pharmacy and Biotechnology, University of Bologna, Bologna, Italy.
Studio Oculistico d'Azeglio, Bologna, Italy.
Transl Vis Sci Technol. 2025 Jun 2;14(6):20. doi: 10.1167/tvst.14.6.20.
Identification and quantification of characteristic visual field (VF) patterns in patients with dominant optic atrophy (DOA) using the archetypal analysis (AA) machine learning algorithm.
In this retrospective study, we collected 30-2 or 24-2 VFs performed with Humphrey Visual Field analyzer from 144 patients (280 eyes) affected by molecularly confirmed DOA carrying OPA1 heterozygous mutation. The VFs were randomly separated into a training set (224 VFs, 80%) and test set (56 VFs, 20%). An AA model was developed by decomposing the VFs of the training set into archetypes (ATs). Spearman correlations were calculated between ATs' weights and mean deviation (MD) and visual acuity (VA). Statistical comparison was performed between ATs weights according to mutation subtype groups.
The DOA-AA model was composed of eight ATs with a high performance in the test set (R2 = 0.88). According to the Ocular Hypertension Treatment Study (OHTS) classification, the central/ceco-central scotoma resembling ATs presented the highest weights (24%) followed by superior defects (13%). ATs with more abnormal VF resembling defects correlated most with MD (AT5-8), whereas only the total loss AT7 with VA (P value < 0.01). Subtype mutations linked with worse clinical features had statistically significantly higher weights for worse ATs (AT7, P < 0.001).
The developed AA model allowed the identification and quantification of VF patterns in DOA. Furthermore, a clinical genotype-phenotype association was supported by the comparison of severity at VF AA decomposition.
AA enables an objective identification of quantifiable visual field defects intrinsic to DOA providing functional details based on genotype.
使用原型分析（AA）机器学习算法识别并量化显性遗传性视神经萎缩（DOA）患者的特征性视野（VF）模式。
在这项回顾性研究中，我们收集了144例经分子确诊携带OPA1杂合突变的DOA患者（280只眼）使用Humphrey视野分析仪进行的30-2或24-2视野检查结果。视野检查结果被随机分为训练集（224个视野，80%）和测试集（56个视野，20%）。通过将训练集的视野分解为原型（AT）来建立AA模型。计算原型权重与平均偏差（MD）和视力（VA）之间的Spearman相关性。根据突变亚型组对原型权重进行统计学比较。
DOA-AA模型由八个原型组成，在测试集中表现出高性能（R2 = 0.88）。根据高眼压治疗研究（OHTS）分类，类似于原型的中心/旁中心暗点权重最高（24%），其次是上方缺损（13%）。视野更异常类似于缺损的原型与MD相关性最高（AT5-8），而只有完全缺损的AT7与VA相关（P值<0.01）。与更差临床特征相关的亚型突变在更差的原型（AT7，P < 0.001）中权重具有统计学显著更高。
所建立的AA模型能够识别并量化DOA中的视野模式。此外，通过视野AA分解时严重程度的比较支持了临床基因型-表型关联。
AA能够客观识别DOA固有的可量化视野缺损，并基于基因型提供功能细节。
