Expert tumor annotations and radiomic features for the ISPY1/ACRIN 6657 trial data collection

This dataset enhances the ISPY1 data collection, with uniformly curated data, tumor annotations, and quantitative imaging features. This dataset includes a) uniformly processed scans that are harmonized to match the intensity and spatial characteristics, facilitating immediate use in computational studies, b) computationally-generated and manually-revised expert annotations of tumor regions, as well as c) a comprehensive set of quantitative imaging (also known as radiomic) features corresponding to the tumor regions.

The segmentations for the ISPY1/ACRIN 6657 dataset currently hosted on TCIA’s website describe a) the tumor volume of interest (VOI) and b) functional tumor volume (FTV).

The provided tumor VOI is a 3D rectangular box enclosing the enhancing tumor region, while including peritumoral tissue. The VOI provides a general guideline of where the tumor is located within patient anatomy, but it does not delineate tumor boundaries or shape.
The FTV segmentations describe only enhancing voxels in the tumor, i.e., defined by peak enhancement or signal enhancement ratio criteria.
These currently provided segmentations do not include non-enhancing portions of the tumor volume, which represent a significant portion of the disease burden that needs to be studied to better understand and quantify the disease.

The segmentations in these new analysis results are for the entire 3D primary lesion, including both the enhancing and the non-enhancing tumor regions, therefore defining the structural tumor volume (STV). These STV annotations were generated by manually delineating the primary lesion volume, after confirming the location of the primary lesion from the provided VOI and FTV. The STV annotations were reviewed and approved by a board-certified, fellowship-trained breast radiologist, and are statistically significantly different from FTV.

We believe these STV annotations will allow analyses of the entire disease burden and analyses of tumor heterogeneity regarding contrast uptake, contributing to further expanding our mechanistic understanding of the disease potentially leading to improved patient management.

本数据集对ISPY1数据集（ISPY1 data collection）进行了拓展优化，包含统一规整质控的影像数据、肿瘤标注信息与定量影像特征。具体包括：a) 经过统一预处理的影像扫描数据已完成强度与空间特征的标准化协调，可直接应用于计算研究；b) 由计算生成并经专家手动修正的肿瘤区域标注；c) 针对肿瘤区域的全套定量影像特征（亦称放射组学特征）。

目前托管于TCIA网站（TCIA’s website）的ISPY1/ACRIN 6657数据集的分割标注包含两类：其一为肿瘤感兴趣体积（tumor volume of interest, VOI），其二为功能性肿瘤体积（functional tumor volume, FTV）。

本次提供的肿瘤VOI为包裹强化肿瘤区域并包含瘤周组织的三维矩形框。该VOI仅可作为肿瘤在患者解剖结构中位置的大致参考，无法勾勒肿瘤的边界与形态。FTV分割标注仅涵盖肿瘤内的强化体素，即基于峰值强化或信号强化比标准定义的区域。

当前提供的上述分割标注未包含肿瘤体积内的非强化区域，而这类区域占疾病负荷的相当比例，需对其开展研究以更好地理解并量化该疾病。

本次新分析结果中的分割标注针对整个三维原发灶，涵盖强化与非强化两类肿瘤区域，由此定义了结构性肿瘤体积（STV, structural tumor volume）。这些STV标注通过手动勾勒原发灶体积生成，生成前已基于提供的VOI与FTV确认了原发灶的位置。该STV标注已由获得专科认证且接受过专科进修培训的乳腺放射科医师审核确认，且与FTV存在统计学意义上的显著差异。

我们认为，这些STV标注可支持针对整体疾病负荷的分析，以及基于造影剂摄取情况的肿瘤异质性分析，有助于进一步深化对该疾病的机制性理解，最终有望改善患者诊疗方案。