Conte Gian Marco, Moassefi Mana, Decker Paul A, Kosel Matthew L, McCarthy Christina B, Sagen Jessica A, Nikanpour Yalda, Fereidan-Esfahani Mahboubeh, Ruff Michael W, Guido Fiorella S, Pump Heather K, Burns Terry C, Jenkins Robert B, Erickson Bradley J, Lachance Daniel H, Tobin W Oliver, Eckel-Passow Jeanette E
From the Department of Radiology (G.M.C., M.M., Y.N., B.J.E.), Mayo Clinic, Rochester, Minnesota.
Deptartment of Quantitative Health Sciences (P.A.D., M.L.K., J.E.E.-P.), Mayo Clinic, Rochester, Minnesota.
AJNR Am J Neuroradiol. 2025 Jul 1;46(7):1412-1420. doi: 10.3174/ajnr.A8645.
Diagnosis of tumefactive demyelination can be challenging. The diagnosis of indeterminate brain lesions on MRI often requires tissue confirmation via brain biopsy. Noninvasive methods for accurate diagnosis of tumor and nontumor etiologies allows for tailored therapy, optimal tumor control, and a reduced risk of iatrogenic morbidity and mortality. Tumefactive demyelination has imaging features that mimic wild-type glioblastoma (wt GBM). We hypothesized that deep learning applied to postcontrast T1-weighted (T1C) and T2-weighted (T2) MRI can discriminate tumefactive demyelination from wt GBM.
Patients with tumefactive demyelination ( = 144) and wt GBM ( = 455) were identified by clinical registries. A 3D DenseNet121 architecture was used to develop models to differentiate tumefactive demyelination and wt GBM by using both T1C and T2 MRI, as well as only T1C and only T2 images. A 3-stage design was used: 1) model development and internal validation via 5-fold cross validation by using a sex-, age-, and MRI technology-matched set of tumefactive demyelination and wt GBM, 2) validation of model specificity on independent wt GBM, and 3) prospective validation on tumefactive demyelination and wt GBM. Stratified area under the receiver operating curves (AUROCs) were used to evaluate model performance stratified by sex, age at diagnosis, MRI scanner strength, and MRI acquisition.
The deep learning model developed by using both T1C and T2 images had a prospective validation AUROC of 88% (95% CI: 0.82-0.95). In the prospective validation stage, a model score threshold of 0.28 resulted in 91% sensitivity of correctly classifying tumefactive demyelination and 80% specificity (correctly classifying wt GBM). Stratified AUROCs demonstrated that model performance may be improved if thresholds were chosen stratified by age and MRI acquisition.
MRI can provide the basis for applying deep learning models to aid in the differential diagnosis of brain lesions. Further validation is needed to evaluate how well the model generalizes across institutions, patient populations, and technology, and to evaluate optimal thresholds for classification. Next steps also should incorporate additional tumor etiologies such as CNS lymphoma and brain metastases.
瘤样脱髓鞘病变的诊断具有挑战性。MRI上不确定脑病变的诊断通常需要通过脑活检进行组织确认。准确诊断肿瘤性和非肿瘤性病因的非侵入性方法有助于制定个性化治疗方案、实现最佳肿瘤控制，并降低医源性发病和死亡风险。瘤样脱髓鞘病变具有类似于野生型胶质母细胞瘤（wt GBM）的影像学特征。我们假设应用于增强后T1加权（T1C）和T2加权（T2）MRI的深度学习能够区分瘤样脱髓鞘病变和wt GBM。
通过临床登记识别出瘤样脱髓鞘病变患者（n = 144）和wt GBM患者（n = 455）。使用3D DenseNet121架构开发模型，通过同时使用T1C和T2 MRI以及仅使用T1C和仅使用T2图像来区分瘤样脱髓鞘病变和wt GBM。采用三阶段设计：1）通过使用性别、年龄和MRI技术匹配的瘤样脱髓鞘病变和wt GBM数据集进行5折交叉验证来开发模型并进行内部验证，2）在独立的wt GBM上验证模型特异性，3）对瘤样脱髓鞘病变和wt GBM进行前瞻性验证。采用分层受试者操作曲线下面积（AUROC）来评估按性别、诊断时年龄、MRI扫描仪强度和MRI采集分层的模型性能。
使用T1C和T2图像开发的深度学习模型的前瞻性验证AUROC为88%（95% CI：0.82 - 0.95）。在前瞻性验证阶段，模型评分阈值为0.28时，正确分类瘤样脱髓鞘病变的敏感性为91%，特异性为80%（正确分类wt GBM）。分层AUROC表明，如果按年龄和MRI采集选择阈值，模型性能可能会提高。
MRI可为应用深度学习模型辅助脑病变的鉴别诊断提供依据。需要进一步验证以评估模型在不同机构、患者群体和技术中的泛化程度，以及评估分类的最佳阈值。下一步还应纳入其他肿瘤病因，如中枢神经系统淋巴瘤和脑转移瘤。
