Ferro Matteo, Tataru Octavian Sabin, Carrieri Giuseppe, Busetto Gian Maria, Falagario Ugo Giovanni, Maggi Martina, Crocetto Felice, Barone Biagio, Del Giudice Francesco, Marchioni Michele, Terracciano Daniela, Lucarelli Giuseppe, Ditonno Pasquale, Gherasim Raul, Todea-Moga Ciprian, Fallara Giuseppe, Tozzi Marco, Cioffi Antonio, Bianchi Roberto, Digiacomo Alessio, Veccia Alessandro, Antonelli Alessandro, Sighinolfi Maria Chiara, Schips Luigi, Rocco Bernardo
Unit of Urology, Department of Health Science, ASST Santi Paolo and Carlo, University of Milan, Milan, Italy.
I.O.S.U.D., George Emil Palade University of Medicine, Pharmacy, Sciences, and Technology of Târgu Mureș, Gheorghe Marinescu Street, 38, Târgu Mureș 540142, Romania.
Ther Adv Urol. 2025 Aug 2;17:17562872251352553. doi: 10.1177/17562872251352553. eCollection 2025 Jan-Dec.
Adrenal lesions, often incidentally detected, present diagnostic challenges in distinguishing benign from malignant or hormonally active lesions. Conventional imaging (computed tomography/magnetic resonance imaging (CT/MRI)) has limitations, driving interest in artificial intelligence (AI) and radiomics for enhanced accuracy.
To systematically evaluate AI and radiomics applications in adrenal lesion characterization, focusing on diagnostic performance, methodologies, and clinical utility.
PRISMA-guided systematic review of studies published up to June 2024.
PubMed, Scopus, Web of Science, and Google Scholar were searched using the keywords: , and . Inclusion followed PICO criteria: patients with indeterminate lesions, AI/radiomics interventions, comparisons to standard diagnostics, and diagnostic accuracy. Two reviewers screened studies, resolving discrepancies via consensus. Eleven retrospective studies (996 patients) met eligibility.
CT-based radiomics (eight studies) achieved a mean AUC of 0.88 (range: 0.84-0.94) in differentiating benign/malignant or functional/non-functional lesions. Top-performing models identified aldosterone-producing adenomas (AUC: 0.99). MRI-based radiomics (three studies) yielded mean AUC: 0.82 (0.72-0.92), with test-set performance declines (e.g., AUC: 0.72) suggesting overfitting. Nuclear medicine (four studies) demonstrated that hybrid 18F-FDG PET/CT models (SUVmax + texture features) achieved an AUC of 0.97 for metastatic versus benign lesions. AI applications extended to intraoperative navigation (AUC: 0.93) and prognostic prediction.
CT-based radiomics outperformed MRI, aligning with guidelines favoring CT for adrenal assessment. AI-enhanced models show promise in refining diagnostics and reducing invasive procedures. However, retrospective designs, small cohorts, and protocol variability limit generalizability. Future work requires multicenter collaboration, standardized protocols, and prospective validation to translate AI/radiomics into clinical practice.
肾上腺病变常为偶然发现，在区分良性与恶性或激素活性病变方面存在诊断挑战。传统成像（计算机断层扫描/磁共振成像（CT/MRI））存在局限性，这激发了人们对人工智能（AI）和放射组学的兴趣，以提高诊断准确性。
系统评价AI和放射组学在肾上腺病变特征描述中的应用，重点关注诊断性能、方法和临床实用性。
按照PRISMA指南对截至2024年6月发表的研究进行系统评价。
使用关键词在PubMed、Scopus、Web of Science和谷歌学术上进行检索。纳入标准遵循PICO标准：病变不确定的患者、AI/放射组学干预、与标准诊断方法的比较以及诊断准确性。两名研究者筛选研究，通过共识解决分歧。11项回顾性研究（996例患者）符合纳入标准。
基于CT的放射组学（8项研究）在区分良性/恶性或功能性/非功能性病变方面的平均曲线下面积（AUC）为0.88（范围：0.84 - 0.94）。表现最佳的模型可识别醛固酮瘤（AUC：0.99）。基于MRI的放射组学（3项研究）的平均AUC为0.82（0.72 - 0.92），测试集性能下降（例如AUC：0.72）表明存在过拟合现象。核医学（4项研究）表明，混合18F - FDG PET/CT模型（SUVmax + 纹理特征）在区分转移性与良性病变方面的AUC为0.97。AI应用扩展到术中导航（AUC：0.93）和预后预测。
基于CT的放射组学优于MRI，这与支持CT用于肾上腺评估的指南一致。AI增强模型在优化诊断和减少侵入性操作方面显示出前景。然而，回顾性设计、小样本队列和方案变异性限制了可推广性。未来的工作需要多中心合作、标准化方案和前瞻性验证，以便将AI/放射组学转化为临床实践。
