Stroke Diagnostic and Monitoring Division, AtheroPoint™, Roseville, CA, USA.
Advanced Knowledge Engineering Centre, GBTI, Roseville, CA, USA.
J Med Syst. 2022 Aug 21;46(10):62. doi: 10.1007/s10916-022-01850-y.
Variations in COVID-19 lesions such as glass ground opacities (GGO), consolidations, and crazy paving can compromise the ability of solo-deep learning (SDL) or hybrid-deep learning (HDL) artificial intelligence (AI) models in predicting automated COVID-19 lung segmentation in Computed Tomography (CT) from unseen data leading to poor clinical manifestations. As the first study of its kind, "COVLIAS 1.0-Unseen" proves two hypotheses, (i) contrast adjustment is vital for AI, and (ii) HDL is superior to SDL. In a multicenter study, 10,000 CT slices were collected from 72 Italian (ITA) patients with low-GGO, and 80 Croatian (CRO) patients with high-GGO. Hounsfield Units (HU) were automatically adjusted to train the AI models and predict from test data, leading to four combinations-two Unseen sets: (i) train-CRO:test-ITA, (ii) train-ITA:test-CRO, and two Seen sets: (iii) train-CRO:test-CRO, (iv) train-ITA:test-ITA. COVILAS used three SDL models: PSPNet, SegNet, UNet and six HDL models: VGG-PSPNet, VGG-SegNet, VGG-UNet, ResNet-PSPNet, ResNet-SegNet, and ResNet-UNet. Two trained, blinded senior radiologists conducted ground truth annotations. Five types of performance metrics were used to validate COVLIAS 1.0-Unseen which was further benchmarked against MedSeg, an open-source web-based system. After HU adjustment for DS and JI, HDL (Unseen AI) > SDL (Unseen AI) by 4% and 5%, respectively. For CC, HDL (Unseen AI) > SDL (Unseen AI) by 6%. The COVLIAS-MedSeg difference was < 5%, meeting regulatory guidelines.Unseen AI was successfully demonstrated using automated HU adjustment. HDL was found to be superior to SDL.
COVID-19 病变（如玻璃样混浊（GGO）、实变和铺路石征）的变化可能会影响基于深度学习的人工智能（AI）模型预测 CT 中 COVID-19 肺分割的能力，这可能导致模型无法识别未见过的数据，从而出现不良的临床表现。作为此类研究的首例，“COVLIAS 1.0-Unseen”验证了两个假设，（i）对比度调整对 AI 至关重要，（ii）HDL 优于 SDL。在一项多中心研究中，从 72 名意大利（ITA）低 GGO 患者和 80 名克罗地亚（CRO）高 GGO 患者中收集了 10000 张 CT 切片。HU 被自动调整以训练 AI 模型并对测试数据进行预测，从而产生了四个组合-两个未见过的数据集：（i）train-CRO:test-ITA，（ii）train-ITA:test-CRO，和两个见过的数据集：（iii）train-CRO:test-CRO，（iv）train-ITA:test-ITA。COVLIAS 使用了三个 SDL 模型：PSPNet、SegNet 和 UNet，以及六个 HDL 模型：VGG-PSPNet、VGG-SegNet、VGG-UNet、ResNet-PSPNet、ResNet-SegNet 和 ResNet-UNet。两名受过训练的、盲目的资深放射科医生进行了地面实况注释。使用了五种性能指标来验证 COVLIAS 1.0-Unseen，该模型还与 MedSeg（一个开源的基于网络的系统）进行了基准测试。在对 DS 和 JI 进行 HU 调整后，HDL（未见过的 AI）分别比 SDL（未见过的 AI）高出 4%和 5%。对于 CC，HDL（未见过的 AI）比 SDL（未见过的 AI）高出 6%。COVLIAS-MedSeg 的差异<5%，符合监管指南。通过自动 HU 调整成功地展示了未见过的 AI。结果发现 HDL 优于 SDL。
