Artificial intelligence (AI) technologies are transforming the way medical images are analyzed, offering unprecedented capabilities in quantitatively extracting features that go beyond traditional visual limitations. These advanced AI techniques, including machine learning and deep learning algorithms, can systematically process complex imaging data, detecting subtle patterns that may be missed by human radiologists.

人工智能（AI）技术正在改变医学图像分析的方式，提供了超越传统视觉限制的定量特征提取能力。这些先进的AI技术，包括机器学习和深度学习算法，能够系统地处理复杂的成像数据，检测到人类放射科医生可能忽略的细微模式。

Now, a recent meta-analysis of 19 studies has highlighted the potential of combining biparametric magnetic resonance imaging (bpMRI) with AI for the robust detection of clinically significant prostate cancer (csPCa)..

现在，一项最近的包含19项研究的荟萃分析强调了将双参数磁共振成像 (bpMRI) 与人工智能结合在可靠检测临床显著性前列腺癌 (csPCa) 方面的潜力。

This meta-analysis, published in

这项荟萃分析，发表于

Academic Radiology

学术放射学

, was performed by researchers from Zhejiang Provincial People’s Hospital (Hangzhou, China) who reviewed data from 6,286 patients. The analysis included 4,594 patients from internal validation cohorts, 795 from external validation cohorts, and 897 patients whose scans were interpreted without AI assistance by radiologists.

，由浙江省级人民医院（中国杭州）的研究人员进行，他们审查了6286名患者的数据。分析包括4594名来自内部验证队列的患者，795名来自外部验证队列的患者，以及897名由放射科医生在没有人工智能辅助的情况下解读扫描结果的患者。

The internal validation cohorts demonstrated an average sensitivity of 88% and specificity of 79%. For the external validation studies, the average sensitivity and specificity were 85% and 83%, respectively, as noted by the authors of the meta-analysis..

内部验证队列表现出平均88%的敏感性和79%的特异性。对于外部验证研究，荟萃分析的作者指出，平均敏感性和特异性分别为85%和83%。

The study also found that the combination of bpMRI and AI achieved a 91% average area under the receiver operating characteristic curve (AUC) for detecting csPCa in both internal and external validation cohorts. This was a significant improvement over the 78% AUC for radiologists interpreting bpMRI without AI assistance.

该研究还发现，bpMRI与AI相结合在内部和外部验证队列中检测csPCa的受试者工作特征曲线下平均面积（AUC）达到了91%，相比没有AI辅助的情况下放射科医生解读bpMRI的78% AUC有了显著提高。

While bpMRI offers several advantages over multiparametric MRI (mpMRI), including shorter exam durations, cost-effectiveness, and better patient safety, the meta-analysis authors noted that its effectiveness could be hindered by morphological constraints and subjective interpretations. However, they emphasized that the integration of deep learning and machine learning could greatly enhance the ability of bpMRI to accurately characterize clinically significant prostate cancer..

虽然bpMRI相比多参数MRI（mpMRI）具有多种优势，包括更短的检查时间、成本效益以及更好的患者安全性，但荟萃分析作者指出，其效果可能受到形态学限制和主观解读的影响。然而，他们强调，深度学习和机器学习的结合可以显著提升bpMRI准确表征具有临床意义的前列腺癌的能力。

“AI improves the accuracy and reliability of tumor classification by effectively extracting morphological features pertinent to PCa,” noted lead study author Guangzhao Yan, M.D. “Moreover, AI reduces the variability associated with the subjective interpretations of radiologists in conventional diagnostic practices, thus providing more objective and consistent analytical results.”.

“AI通过有效提取与PCa相关的关键形态学特征，提高了肿瘤分类的准确性和可靠性，”该研究的主要作者颜广钊博士指出。“此外，AI减少了传统诊断实践中放射科医生主观解读所带来的差异性，从而提供更客观和一致的分析结果。”

Visit expo >

访问博览会 >

LED-Based X-Ray Viewer

基于LED的X射线观察器

Dixion X-View

迪克森 X-View

New

新建

Ultrasonic Pocket Doppler

超声波口袋多普勒

SD1

SD1

Digital X-Ray Detector Panel

数字X射线探测器面板

Acuity G4

锐度 G4

3T MRI Scanner

3T核磁共振扫描仪

MAGNETOM Cima.X

磁共振 Cima.X