本研究的目的是研究扩展的自适应 nnU-Net 框架，用于在磁共振成像 (MRI) 上检测和分割脑转移瘤 (BM)。六种不同的 nnU-Net 系统，具有自适应数据采样、自适应 Dice 损失或使用来自 7 个机构的 2092 名患者（1712、195 和 185 名患者）对不同的贴片/批次大小进行了训练和测试，用于在 3 维（3D）后 Gd T1 加权 MRI 体积上检测和分割尺寸≥2 mm 的实质内 BM。分别是训练、验证和测试）。回顾性收集由医生为立体定向放射外科描绘的 BM 肿瘤体积，并在每个研究所进行整理。 2 名放射科医生进行了额外的集中数据管理，以创建未轮廓 BM 的总肿瘤体积，以提高地面实况的准确性。使用 3D 生成管道通过 1025 个 MRI 体积的合成 BM 增强了训练数据集。 BM 检测通过病变水平敏感性和假阳性 (FP) 率进行评估。 BM 分割通过病变级别 Dice 相似系数、95 百分位数 Hausdorff 距离和平均 Hausdorff 距离 (HD) 进行评估。性能是根据不同的 BM 大小进行评估的。使用 206 名患者的第二个数据集进行了额外的测试。在 6 个 nnU-Net 系统中，具有自适应 Dice 损失的 nnU-Net 在第一个测试数据集上实现了最佳检测和分割性能。在 FP 率为 0.65 ± 1.17 时，所有大小的 BM 的总体灵敏度为 0.904，BM ≥0.1 cm3 的总体灵敏度为 0.966，BM <0.1 cm3 的总体灵敏度为 0.824。所有检测到的BM的Dice相似系数、95百分位数豪斯多夫距离和平均HD的平均值分别为0.758、1.45和0.23毫米。第二个测试数据集的性能在所有 BM 尺寸的 FP 速率为 0.57 ± 0.85 的情况下实现了 0.907 的灵敏度，并且所有检测到的 BM 的平均 HD 为 0.33 mm。我们提出的自配置 nnU-Net 框架的扩展实质上提高了小 BM 检测灵敏度，同时保持受控的 FP 率。将研究扩展的 nnU-Net 模型在协助早期 BM 检测和立体定向放射外科计划方面的临床实用性。版权所有 © 2024 Elsevier Inc. 保留所有权利。

The purpose of this study was to investigate an extended self-adapting nnU-Net framework for detecting and segmenting brain metastases (BM) on magnetic resonance imaging (MRI).Six different nnU-Net systems with adaptive data sampling, adaptive Dice loss, or different patch/batch sizes were trained and tested for detecting and segmenting intraparenchymal BM with a size ≥2 mm on 3 Dimensional (3D) post-Gd T1-weighted MRI volumes using 2092 patients from 7 institutions (1712, 195, and 185 patients for training, validation, and testing, respectively). Gross tumor volumes of BM delineated by physicians for stereotactic radiosurgery were collected retrospectively and curated at each institute. Additional centralized data curation was carried out to create gross tumor volumes of uncontoured BM by 2 radiologists to improve the accuracy of ground truth. The training data set was augmented with synthetic BMs of 1025 MRI volumes using a 3D generative pipeline. BM detection was evaluated by lesion-level sensitivity and false-positive (FP) rate. BM segmentation was assessed by lesion-level Dice similarity coefficient, 95-percentile Hausdorff distance, and average Hausdorff distance (HD). The performances were assessed across different BM sizes. Additional testing was performed using a second data set of 206 patients.Of the 6 nnU-Net systems, the nnU-Net with adaptive Dice loss achieved the best detection and segmentation performance on the first testing data set. At an FP rate of 0.65 ± 1.17, overall sensitivity was 0.904 for all sizes of BM, 0.966 for BM ≥0.1 cm3, and 0.824 for BM <0.1 cm3. Mean values of Dice similarity coefficient, 95-percentile Hausdorff distance, and average HD of all detected BMs were 0.758, 1.45, and 0.23 mm, respectively. Performances on the second testing data set achieved a sensitivity of 0.907 at an FP rate of 0.57 ± 0.85 for all BM sizes, and an average HD of 0.33 mm for all detected BM.Our proposed extension of the self-configuring nnU-Net framework substantially improved small BM detection sensitivity while maintaining a controlled FP rate. Clinical utility of the extended nnU-Net model for assisting early BM detection and stereotactic radiosurgery planning will be investigated.Copyright © 2024 Elsevier Inc. All rights reserved.