In The Lancet. Digital health

Background : Prompt identification of patients suspected to have COVID-19 is crucial for disease control. We aimed to develop a deep learning algorithm on the basis of chest CT for rapid triaging in fever clinics.

Methods : We trained a U-Net-based model on unenhanced chest CT scans obtained from 2447 patients admitted to Tongji Hospital (Wuhan, China) between Feb 1, 2020, and March 3, 2020 (1647 patients with RT-PCR-confirmed COVID-19 and 800 patients without COVID-19) to segment lung opacities and alert cases with COVID-19 imaging manifestations. The ability of artificial intelligence (AI) to triage patients suspected to have COVID-19 was assessed in a large external validation set, which included 2120 retrospectively collected consecutive cases from three fever clinics inside and outside the epidemic centre of Wuhan (Tianyou Hospital [Wuhan, China; area of high COVID-19 prevalence], Xianning Central Hospital [Xianning, China; area of medium COVID-19 prevalence], and The Second Xiangya Hospital [Changsha, China; area of low COVID-19 prevalence]) between Jan 22, 2020, and Feb 14, 2020. To validate the sensitivity of the algorithm in a larger sample of patients with COVID-19, we also included 761 chest CT scans from 722 patients with RT-PCR-confirmed COVID-19 treated in a makeshift hospital (Guanggu Fangcang Hospital, Wuhan, China) between Feb 21, 2020, and March 6, 2020. Additionally, the accuracy of AI was compared with a radiologist panel for the identification of lesion burden increase on pairs of CT scans obtained from 100 patients with COVID-19.

Findings : In the external validation set, using radiological reports as the reference standard, AI-aided triage achieved an area under the curve of 0·953 (95% CI 0·949-0·959), with a sensitivity of 0·923 (95% CI 0·914-0·932), specificity of 0·851 (0·842-0·860), a positive predictive value of 0·790 (0·777-0·803), and a negative predictive value of 0·948 (0·941-0·954). AI took a median of 0·55 min (IQR: 0·43-0·63) to flag a positive case, whereas radiologists took a median of 16·21 min (11·67-25·71) to draft a report and 23·06 min (15·67-39·20) to release a report. With regard to the identification of increases in lesion burden, AI achieved a sensitivity of 0·962 (95% CI 0·947-1·000) and a specificity of 0·875 (95 %CI 0·833-0·923). The agreement between AI and the radiologist panel was high (Cohen's kappa coefficient 0·839, 95% CI 0·718-0·940).

Interpretation : A deep learning algorithm for triaging patients with suspected COVID-19 at fever clinics was developed and externally validated. Given its high accuracy across populations with varied COVID-19 prevalence, integration of this system into the standard clinical workflow could expedite identification of chest CT scans with imaging indications of COVID-19.

Funding : Special Project for Emergency of the Science and Technology Department of Hubei Province, China.

Wang Minghuan, Xia Chen, Huang Lu, Xu Shabei, Qin Chuan, Liu Jun, Cao Ying, Yu Pengxin, Zhu Tingting, Zhu Hui, Wu Chaonan, Zhang Rongguo, Chen Xiangyu, Wang Jianming, Du Guang, Zhang Chen, Wang Shaokang, Chen Kuan, Liu Zheng, Xia Liming, Wang Wei

2020-Oct

In International journal of medical informatics ; h5-index 49.0

BACKGROUND : With the recent advances in computational science, machine-learning methods have been increasingly used in medical research. Because such projects usually require both a clinician and a computational data scientist, there is a need for interdisciplinary research collaboration. However, there has been no published analysis of research collaboration networks in cardiovascular medicine using machine intelligence.

METHODS : Co-authorship network analysis was conducted on 2857 research articles published between 2009 and 2019. Bibliographic data were collected from the Web of Science, and the co-authorship networks were represented as undirected multigraphs. The network density, average degree, clustering coefficient, and number of communities were calculated, and the chronological changes were assessed. Thereafter, the leading authors were identified according to the centrality metrics. Finally, we investigated the significance of the characteristics of the co-authorship network in the largest component via a Monte Carlo simulation with the Barabasi-Albert model.

RESULTS : The co-authorship network of the entire period consisted of 13,979 nodes and 68,668 weighted edges. A time-series analysis revealed a linear correlation between the number of nodes and the number of edges (R² = 0.9937, p < 0.001). Additionally, the number of communities was linearly correlated with the number of nodes (R² = 0.9788, p < 0.001). The average shortest path increased by a greater degree than the logarithm of the number of nodes, indicating the scale-free structure of the network. We identified D. Berman as the most central author with regard to the degree centrality and closeness centrality. S. Neubauer was the top-ranking author with regard to the betweenness centrality. Among the 22 authors who were ranked in the top 10 for any centrality, 14 authors (63.6%) had a medical degree (medical doctor, MD). The remaining eight non-MD researchers had a PhD in computational science-related fields. The number of communities detected in the Barabasi-Albert model simulation was similar to that for the largest component of the real network (6.21 ± 0.07 vs. 6, p = 0.096).

CONCLUSIONS : A co-authorship network analysis revealed a structure of collaboration networks in the application of machine learning in the field of cardiovascular disease, which can be useful for planning future scientific collaboration.

Higaki Akinori, Uetani Teruyoshi, Ikeda Shuntaro, Yamaguchi Osamu

2020-Sep-19

Cardiovascular research, Co-authorship network analysis, Machine learning, Natural language processing, Social network analysis

In World journal of gastroenterology ; h5-index 103.0

BACKGROUND : Efforts should be made to develop a deep-learning diagnosis system to distinguish pancreatic cancer from benign tissue due to the high morbidity of pancreatic cancer.

AIM : To identify pancreatic cancer in computed tomography (CT) images automatically by constructing a convolutional neural network (CNN) classifier.

METHODS : A CNN model was constructed using a dataset of 3494 CT images obtained from 222 patients with pathologically confirmed pancreatic cancer and 3751 CT images from 190 patients with normal pancreas from June 2017 to June 2018. We established three datasets from these images according to the image phases, evaluated the approach in terms of binary classification (i.e., cancer or not) and ternary classification (i.e., no cancer, cancer at tail/body, cancer at head/neck of the pancreas) using 10-fold cross validation, and measured the effectiveness of the model with regard to the accuracy, sensitivity, and specificity.

RESULTS : The overall diagnostic accuracy of the trained binary classifier was 95.47%, 95.76%, 95.15% on the plain scan, arterial phase, and venous phase, respectively. The sensitivity was 91.58%, 94.08%, 92.28% on three phases, with no significant differences (χ² = 0.914, P = 0.633). Considering that the plain phase had same sensitivity, easier access, and lower radiation compared with arterial phase and venous phase , it is more sufficient for the binary classifier. Its accuracy on plain scans was 95.47%, sensitivity was 91.58%, and specificity was 98.27%. The CNN and board-certified gastroenterologists achieved higher accuracies than trainees on plain scan diagnosis (χ² = 21.534, P < 0.001; χ² = 9.524, P < 0.05; respectively). However, the difference between CNN and gastroenterologists was not significant (χ² = 0.759, P = 0.384). In the trained ternary classifier, the overall diagnostic accuracy of the ternary classifier CNN was 82.06%, 79.06%, and 78.80% on plain phase, arterial phase, and venous phase, respectively. The sensitivity scores for detecting cancers in the tail were 52.51%, 41.10% and, 36.03%, while sensitivity for cancers in the head was 46.21%, 85.24% and 72.87% on three phases, respectively. Difference in sensitivity for cancers in the head among the three phases was significant (χ² = 16.651, P < 0.001), with arterial phase having the highest sensitivity.

CONCLUSION : We proposed a deep learning-based pancreatic cancer classifier trained on medium-sized datasets of CT images. It was suitable for screening purposes in pancreatic cancer detection.

Ma Han, Liu Zhong-Xin, Zhang Jing-Jing, Wu Feng-Tian, Xu Cheng-Fu, Shen Zhe, Yu Chao-Hui, Li You-Ming

2020-Sep-14

Computed tomography, Convolutional neural networks, Deep learning, Pancreatic cancer

In The Lancet. Digital health

Zhou Yadi, Wang Fei, Tang Jian, Nussinov Ruth, Cheng Feixiong

2020-Sep-18

In NeuroImage. Clinical

OBJECTIVE : Focal cortical dysplasia (FCD) is the most common epileptogenic developmental malformation and a prevalent cause of surgically amenable epilepsy. While cellular and molecular biology data suggest that FCD lesional characteristics lie along a spectrum, this notion remains to be verified in vivo. We tested the hypothesis that machine learning applied to MRI captures FCD lesional variability at a mesoscopic scale.

METHODS : We studied 46 patients with histologically verified FCD Type II and 35 age- and sex-matched healthy controls. We applied consensus clustering, an unsupervised learning technique that identifies stable clusters based on bootstrap-aggregation, to 3 T multicontrast MRI (T1-weighted MRI and FLAIR) features of FCD normalized with respect to distributions in controls.

RESULTS : Lesions were parcellated into four classes with distinct structural profiles variably expressed within and across patients: Class-1 with isolated white matter (WM) damage; Class-2 combining grey matter (GM) and WM alterations; Class-3 with isolated GM damage; Class-4 with GM-WM interface anomalies. Class membership was replicated in two independent datasets. Classes with GM anomalies impacted local function (resting-state fMRI derived ALFF), while those with abnormal WM affected large-scale connectivity (assessed by degree centrality). Overall, MRI classes reflected typical histopathological FCD characteristics: Class-1 was associated with severe WM gliosis and interface blurring, Class-2 with severe GM dyslamination and moderate WM gliosis, Class-3 with moderate GM gliosis, Class-4 with mild interface blurring. A detection algorithm trained on class-informed data outperformed a class-naïve paradigm.

SIGNIFICANCE : Machine learning applied to widely available MRI contrasts uncovers FCD Type II variability at a mesoscopic scale and identifies tissue classes with distinct structural dimensions, functional and histopathological profiles. Integrating in vivo staging of FCD traits with automated lesion detection is likely to inform the development of novel personalized treatments.

Lee Hyo M, Gill Ravnoor S, Fadaie Fatemeh, Cho Kyoo H, Guiot Marie C, Hong Seok-Jun, Bernasconi Neda, Bernasconi Andrea

2020-Sep-18

Cortical dysplasia, Epilepsy, MRI

In Computer methods and programs in biomedicine

BACKGROUND AND OBJECTIVE : In medical imaging, population studies have to overcome the differences that exist between individuals to identify invariant image features that can be used for diagnosis purposes. In functional neuroimaging, an appealing solution to identify neural coding principles that hold at the population level is inter-subject pattern analysis, i.e. to learn a predictive model on data from multiple subjects and evaluate its generalization performance on new subjects. Although it has gained popularity in recent years, its widespread adoption is still hampered by the blatant lack of a formal definition in the literature. In this paper, we precisely introduce the first principled formalization of inter-subject pattern analysis targeted at multivariate group analysis of functional neuroimaging.

METHODS : We propose to frame inter-subject pattern analysis as a multi-source transductive transfer question, thus grounding it within several well defined machine learning settings and broadening the spectrum of usable algorithms. We describe two sets of inter-subject brain decoding experiments that use several open datasets: a magneto-encephalography study with 16 subjects and a functional magnetic resonance imaging paradigm with 100 subjects. We assess the relevance of our framework by performing model comparisons, where one brain decoding model exploits our formalization while others do not.

RESULTS : The first set of experiments demonstrates the superiority of a brain decoder that uses subject-by-subject standardization compared to state of the art models that use other standardization schemes, making the case for the interest of the transductive and the multi-source components of our formalization The second set of experiments quantitatively shows that, even after such transformation, it is more difficult for a brain decoder to generalize to new participants rather than to new data from participants available in the training phase, thus highlighting the transfer gap that needs to be overcome.

CONCLUSION : This paper describes the first formalization of inter-subject pattern analysis as a multi-source transductive transfer learning problem. We demonstrate the added value of this formalization using proof-of-concept experiments on several complementary functional neuroimaging datasets. This work should contribute to popularize inter-subject pattern analysis for functional neuroimaging population studies and pave the road for future methodological innovations.

Wang Qi, Artières Thierry, Takerkart Sylvain

2020-Sep-11

Functional neuroimaging, Machine learning, Neuroinformatics, Population studies