多发性硬化症中自动病变分割软件的横断面验证：与放射科医生评估的比较

Cross-Sectional Validation of an Automated Lesion Segmentation Software in Multiple Sclerosis: Comparison with Radiologist Assessments.

作者信息

Spampinato Maria Vittoria, Collins Heather R, Wells Hannah, Dennis William, Chamberlin Jordan H, Ye Emily, Chetta Justin A, Matheus Maria Gisele, Stalcup Seth T, Roberts Donna R

机构信息

From the Department of Radiology (M.V.S., H.R.C., W.D., J.H.C, J.A.C., M.G.M., S.T.S., D.R.R.), Medical University of South Carolina, Charleston, South Carolina

From the Department of Radiology (M.V.S., H.R.C., W.D., J.H.C, J.A.C., M.G.M., S.T.S., D.R.R.), Medical University of South Carolina, Charleston, South Carolina.

出版信息

AJNR Am J Neuroradiol. 2025 Jul 1;46(7):1510-1516. doi: 10.3174/ajnr.A8655.

DOI:10.3174/ajnr.A8655

PMID:39805667

Abstract

BACKGROUND AND PURPOSE

MRI is widely used to assess disease burden in MS. This study aimed to evaluate the effectiveness of a commercially available k-nearest neighbors (kNN) network software in quantifying white matter lesion (WML) burden in MS. We compared the software's WML quantification to expert radiologists' assessments.

MATERIALS AND METHODS

We retrospectively reviewed brain MRI examinations of adult patients with MS and of adult patients without MS and with a normal brain MRI referred from the neurology clinic. MR images were processed by using an AI-powered, cloud-based kNN software, which generated a DICOM lesion distribution map and a report of WML count and volume in 4 brain regions (periventricular, deep, juxtacortical, and infratentorial white matter). Two blinded radiologists performed semiquantitative assessments of WM lesion load and lesion segmentation accuracy. Additionally, 4 blinded neuroradiologists independently reviewed the data to determine if MRI findings supported an MS diagnosis. The associations between radiologist-rated WML load and kNN model WML volume and count were evaluated with Spearman rank order correlation coefficient (rho) because these variables were not normally distributed. Results were considered significant when < .05.

RESULTS

The study included 32 patients with MS (35.4 years ±9.1) and 19 patients without MS (33.5 years ±12.1). The kNN software demonstrated 94.1% and 84.3% accuracy in differentiating MS from non-MS subjects based respectively on WML count and WML volume, compared with radiologists' accuracy of 90.2% to 94.1%. Lesion segmentation was more accurate for the deep WM and infratentorial regions than for the juxtacortical region (both < .001).

CONCLUSIONS

kNN-derived WML volume and WML count provide valuable quantitative metrics of disease burden in MS. AI-powered postprocessing software may enhance the interpretation of brain MRIs in MS patients.

摘要

背景与目的

磁共振成像（MRI）被广泛用于评估多发性硬化症（MS）的疾病负担。本研究旨在评估一款市售的k近邻（kNN）网络软件在量化MS患者白质病变（WML）负担方面的有效性。我们将该软件对WML的量化结果与放射科专家的评估进行了比较。

材料与方法

我们回顾性分析了来自神经科门诊的成年MS患者以及成年非MS患者且脑MRI正常者的脑部MRI检查结果。使用一款基于云计算的人工智能kNN软件对MR图像进行处理，该软件生成了一个DICOM病变分布图以及4个脑区（脑室周围、深部、皮质下和幕下白质）的WML计数和体积报告。两名不知情的放射科医生对WM病变负荷和病变分割准确性进行了半定量评估。此外，4名不知情的神经放射科医生独立审查数据，以确定MRI结果是否支持MS诊断。由于这些变量并非正态分布，因此使用Spearman等级相关系数（rho）评估放射科医生评定的WML负荷与kNN模型WML体积和计数之间的相关性。当<0.05时，结果被认为具有统计学意义。

结果

该研究纳入了32例MS患者（35.4岁±9.1）和19例非MS患者（33.5岁±12.1）。基于WML计数和WML体积，kNN软件在区分MS与非MS受试者方面的准确率分别为94.1%和84.3%，而放射科医生的准确率为90.2%至94.1%。深部WM和幕下区域的病变分割比皮质下区域更准确（均<0.001）。