现在的位置: 首页时讯速递, 进展交流>正文
[Lancet最新论文]:深度学习算法检测头颅CT的关键发现
2018年12月05日 时讯速递, 进展交流 暂无评论

Deep learning algorithms for detection of critical findings in head CT scans: a retrospective study

Sasank Chilamkurthy, Rohit Ghosh, Swetha Tanamala, et al

Lancet 2018; 392: 2388-2396

Summary

Background 背景

Non-contrast head CT scan is the current standard for initial imaging of patients with head trauma or stroke symptoms. We aimed to develop and validate a set of deep learning algorithms for automated detection of the following key findings from these scans: intracranial haemorrhage and its types (ie, intraparenchymal, intraventricular, subdural, extradural, and subarachnoid); calvarial fractures; midline shift; and mass effect.

头颅平扫CT是目前对于头颅创伤或卒中患者的标准初始影像学检查。我们旨在建立并验证一种深度学习算法,用于自动检测以下关键发现:脑内出血及其类型(即脑实质内、脑室内、硬膜下、硬膜外及蛛网膜下腔);颅骨骨折;中线移位及占位效应。

Methods 方法

We retrospectively collected a dataset containing 313 318 head CT scans together with their clinical reports from around 20 centres in India between Jan 1, 2011, and June 1, 2017. A randomly selected part of this dataset (Qure25k dataset) was used for validation and the rest was used to develop algorithms. An additional validation dataset (CQ500 dataset) was collected in two batches from centres that were different from those used for the development and Qure25k datasets. We excluded postoperative scans and scans of patients younger than 7 years. The original clinical radiology report and consensus of three independent radiologists were considered as gold standard for the Qure25k and CQ500 datasets, respectively. Areas under the receiver operating characteristic curves (AUCs) were primarily used to assess the algorithms.

我们从印度20个中心回顾性收集了2011年1月1日至2017年6月1日间的313318份头颅CT扫描及其临床报告。从上述数据集中随机选择部分(Qure25k dataset)用于验证,其余影像用于建立算法。从其他中心收集另外的2批影像作为验证数据集(CQ500 dataset)。我们排除了术后CT扫描及7岁以下儿童的扫描结果。原始临床影像报告及3名独立放射科医生的共识分别作为Qure25k dataset和CQ500 dataset的金标准。采用受试者工作特征曲线下面积(AUC)评估算法准确性。

Findings 结果

The Qure25k dataset contained 21 095 scans (mean age 43 years; 9030 [43%] female patients), and the CQ500 dataset consisted of 214 scans in the first batch (mean age 43 years; 94 [44%] female patients) and 277 scans in the second batch (mean age 52 years; 84 [30%] female patients). On the Qure25k dataset, the algorithms achieved an AUC of 0·92 (95% CI 0·91–0·93) for detecting intracranial haemorrhage (0·90 [0·89–0·91] for intraparenchymal, 0·96 [0·94–0·97] for intraventricular, 0·92 [0·90–0·93] for subdural, 0·93 [0·91–0·95] for extradural, and 0·90 [0·89–0·92] for subarachnoid). On the CQ500 dataset, AUC was 0·94 (0·92–0·97) for intracranial haemorrhage (0·95 [0·93–0·98], 0·93 [0·87–1·00], 0·95 [0·91–0·99], 0·97 [0·91–1·00], and 0·96 [0·92–0·99], respectively). AUCs on the Qure25k dataset were 0·92 (0·91–0·94) for calvarial fractures, 0·93 (0·91–0·94) for midline shift, and 0·86 (0·85–0·87) for mass effect, while AUCs on the CQ500 dataset were 0·96 (0·92–1·00), 0·97 (0·94–1·00), and 0·92 (0·89–0·95), respectively.

Qure25k dataset包括 21 095 份扫描(平均年龄43岁;9030名 [43%] 女性患者),CQ500 dataset的第一批影像包括 214 份扫描(平均年龄43岁;94名 [44%] 女性患者),第二批影像包括 277 份扫描(平均年龄52岁;84名 [30%] 女性患者)。采用Qure25k dataset,算法检测脑内出血的AUC 为 0·92 (95% CI 0·91–0·93)(脑实质内出血0·90 [0·89–0·91],脑室内出血 0·96 [0·94–0·97],硬膜下出血 0·92 [0·90–0·93],硬膜外出血 0·93 [0·91–0·95],蛛网膜下腔出血0·90 [0·89–0·92])。采用CQ500 dataset,脑内出血AUC 为 0·94 (0·92–0·97)(分别为0·95 [0·93–0·98], 0·93 [0·87–1·00], 0·95 [0·91–0·99], 0·97 [0·91–1·00], 和 0·96 [0·92–0·99])。Qure25k dataset针对颅骨骨折的AUC为0·92 (0·91–0·94),中线移位0·93 (0·91–0·94),占位效应0·86 (0·85–0·87),而CQ500 dataset的AUC分别为 0·96 (0·92–1·00), 0·97 (0·94–1·00), 和 0·92 (0·89–0·95)。


Interpretation 结论

Our results show that deep learning algorithms can accurately identify head CT scan abnormalities requiring urgent attention, opening up the possibility to use these algorithms to automate the triage process.

我们的研究结果表明,深度学习算法能够准确检测需要紧急关注的头颅CT扫描异常,从而开启了采用这些算法进行自动检诊过程的可能性。

Funding

Qure.ai.

给我留言

您必须 [ 登录 ] 才能发表留言!

×
腾讯微博