Positano Vincenzo, Meloni Antonella, De Santi Lisa Anita, Pistoia Laura, Borsellino Zelia, Cossu Alberto, Massei Francesco, Sanna Paola Maria Grazia, Santarelli Maria Filomena, Cademartiri Filippo
Bioengineering Unit, Fondazione G. Monasterio CNR-Regione Toscana, Pisa, Italy.
Department of Radiology, Fondazione G. Monasterio CNR-Regione Toscana, Pisa, Italy.
Eur Radiol. 2025 Mar;35(3):1522-1532. doi: 10.1007/s00330-024-11245-x. Epub 2024 Dec 10.
To develop a deep-learning model for supervised classification of myocardial iron overload (MIO) from magnitude T2* multi-echo MR images.
Eight hundred twenty-three cardiac magnitude T2* multi-slice, multi-echo MR images from 496 thalassemia major patients (285 females, 57%), labeled for MIO level (normal: T2* > 20 ms, moderate: 10 ≤ T2* ≤ 20 ms, severe: T2* < 10 ms), were retrospectively studied. Two 2D convolutional neural networks (CNN) developed for multi-slice (MS-HippoNet) and single-slice (SS-HippoNet) analysis were trained using 5-fold cross-validation. Performance was assessed using micro-average, multi-class accuracy, and single-class accuracy, sensitivity, and specificity. CNN performance was compared with inter-observer agreement between radiologists on 20% of the patients. The agreement between patients' classifications was assessed by the inter-agreement Kappa test.
Among the 165 images in the test set, a multi-class accuracy of 0.885 and 0.836 was obtained for MS- and SS-Hippo-Net, respectively. Network performances were confirmed on external test set analysis (0.827 and 0.793 multi-class accuracy, 29 patients from the CHMMOTv1 database). The agreement between automatic and ground truth classification was good (MS: κ = 0.771; SS: κ = 0.614), comparable with the inter-observer agreement (MS: κ = 0.872, SS: κ = 0.907) evaluated on the test set.
The developed networks performed classification of MIO level from multiecho, bright-blood, and T2* images with good performances.
Question MRI T2* represents the established clinical tool for MIO assessment. Quality control of the image analysis is a problem in small centers. Findings Deep learning models can perform MIO staging with good accuracy, comparable to inter-observer variability of the standard procedure. Clinical relevance CNN can perform automated staging of cardiac iron overload from multiecho MR sequences facilitating non-invasive evaluation of patients with various hematologic disorders.
开发一种深度学习模型，用于从T2*加权多回波磁共振成像中对心肌铁过载（MIO）进行监督分类。
回顾性研究了496例重型地中海贫血患者（285例女性，占57%）的823幅心脏T2加权多层多回波磁共振图像，这些图像根据MIO水平进行了标注（正常：T2＞20ms，中度：10≤T2*≤20ms，重度：T2*＜10ms）。使用5折交叉验证对为多层分析（MS-HippoNet）和单层分析（SS-HippoNet）开发的两个二维卷积神经网络（CNN）进行训练。使用微平均、多类准确率以及单类准确率、敏感性和特异性来评估性能。将CNN的性能与放射科医生对20%患者的观察者间一致性进行比较。通过一致性Kappa检验评估患者分类之间的一致性。
在测试集中的165幅图像中，MS-HippoNet和SS-HippoNet的多类准确率分别为0.885和0.836。在外部测试集分析（来自CHMMOTv1数据库的29例患者，多类准确率分别为0.827和0.793）中证实了网络性能。自动分类与真实分类之间的一致性良好（MS：κ = 0.771；SS：κ = 0.614），与在测试集上评估的观察者间一致性（MS：κ = 0.872，SS：κ = 0.907）相当。
所开发的网络能够从多回波、亮血和T2*图像中对MIO水平进行分类，性能良好。
问题MRI T2*是评估MIO的既定临床工具。图像分析的质量控制在小中心是个问题。发现深度学习模型能够以良好的准确率进行MIO分期，与标准程序的观察者间变异性相当。临床意义CNN能够从多回波磁共振序列中对心脏铁过载进行自动分期，有助于对各种血液系统疾病患者进行无创评估。
