Given the outbreak of COVID-19 pandemic and the shortage of medical resource, extensive deep learning models have been proposed for automatic COVID-19 diagnosis, based on 3D computed tomography (CT) scans. However, the existing models independently process the 3D lesion segmentation and disease classification, ignoring the inherent correlation between these two tasks. In this paper, we propose a joint deep learning model of 3D lesion segmentation and classification for diagnosing COVID-19, called DeepSC-COVID, as the first attempt in this direction. Specifically, we establish a large-scale CT database containing 1,805 3D CT scans with fine-grained lesion annotations, and reveal 4 findings about lesion difference between COVID-19 and community acquired pneumonia (CAP). Inspired by our findings, DeepSC-COVID is designed with 3 subnets: a cross-task feature subnet for feature extraction, a 3D lesion subnet for lesion segmentation, and a classification subnet for disease diagnosis. Besides, the task-aware loss is proposed for learning the task interaction across the 3D lesion and classification subnets. Different from all existing models for COVID-19 diagnosis, our model is interpretable with fine-grained 3D lesion distribution. Finally, extensive experimental results show that the joint learning framework in our model significantly improves the performance of 3D lesion segmentation and disease classification in both efficiency and efficacy.

A knowledge-augmented deep learning model could accurately segment and detect cerebral aneurysms at CT angiography with no evidence of a significant difference in diagnostic performance compared with clinical radiology reports.

Purpose: Studies have demonstrated that red blood cell distribution width (RDW) is closely associated with the prognosis of patients with chronic obstructive pulmonary disease (COPD). In addition, the dynamic changes in RDW appear to play an important role. Thus, we aimed to investigate the relationship between dynamic changes in RDW and 30-day all-cause readmission of patients with acute exacerbation of COPD (AECOPD). Methods: In this retrospective cohort study, we enrolled patients with AECOPD hospitalized in the Department of Respiratory Medicine in Liyuan Hospital (Wuhan China), a tertiary, university-affiliated, public hospital. Patients with AECOPD were divided into three groups based on their RDW values after the first and fourth days of admission. The normal range for RDW is 10-15%. Patients with normal RDW values were included in the normal group. Patients with an RDW value >15% on the first day, which subsequently decreased by >2% on the fourth day was included in the decreased group. The increased group was comprised of patients with an RDW value >15% on the first day which continued to increase, or those with a normal RDW value on the first day which increased >15% on the fourth day. Results: A total of 239 patients (age: 72 years [range: 64-81 years]; male: n=199 [83.3%]) were included. There were 108, 72, and 59 patients in the RDW normal, decreased, and increased groups, respectively; the 30-day all-cause readmission rate was 9.3%, 9.7%, 27.1%, respectively; (p=0.003), being noticeably higher in the RDW increased group. Dynamic increase of RDW (OR:3.45, 95% CI: 1.39-8.58, p= 0.008) was independently correlated with 30-day all-cause readmission of patients with AECOPD. Conclusion: The dynamic increase of RDW is an independent prognostic factor of 30-day all-cause readmission of patients with AECOPD.

A deep learning model incorporating radiology report–derived textual features outperformed a basic deep learning model trained without textual features in identifying 14 brain diseases and normal brain on MR images while providing interpretable attention maps.

"Just Accepted" papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content. Purpose To develop an MRI-based model for clinically significant prostate cancer (csPCa) diagnosis that can resist rectal artifact interference. Materials and Methods This retrospective study included 2203 male patients with prostate lesions who underwent biparametric MRI and biopsy between January 2019 and June 2023. A targeted adversarial training with proprietary adversarial samples (TPAS) strategy was proposed to enhance model resistance against rectal artifacts. The automated csPCa diagnostic models trained with and without TPAS were compared using multicenter validation datasets. The impact of rectal artifacts on diagnostic performance of each model at the patient and lesion levels was compared using the area under the receiver operating characteristic curve (AUC) and the area under the precision-recall curve (AUPRC). The AUC between models was compared using the Delong test, and the AUPRC was compared using the bootstrap method. Results The TPAS model exhibited diagnostic performance improvements of 6% at the patient level (AUC: 0.87 versus 0.81; P < .001) and 7% at the lesion level (AUPRC: 0.84 versus 0.77; P = .007) compared with the control model. The TPAS model demonstrated less performance decline in the presence of rectal artifact-pattern adversarial noise than the control model (ΔAUC: -17% versus -19%; ΔAUPRC: -18% versus -21%). The TPAS model performed better than the control model in patients with moderate (AUC: 0.79 versus 0.73; AUPRC: 0.68 versus 0.61), and severe (AUC: 0.75 versus 0.57; AUPRC: 0.69 versus 0.59) artifacts. Conclusion This study demonstrates that the TPAS model can reduce rectal artifact interference in MRI-based PCa diagnosis, thereby improving its performance in clinical applications. Published under a CC BY 4.0 license.