PYPei Yang

Pei Yang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Pei Yang has a total of 225 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Large-scale esophageal cancer screening through noncontrast computed tomography and artificial intelligence

    by Jian Zhou, Guangyu Guo, Jiawen Yao, Qinji Yu, Chunpeng Zheng, Xiaochen Feng, Xianghua Ye, Yongjin Zhou, Pei Yang, Lukas Lambert, et al.

    The absence of accurate, noninvasive, scalable screening tools keeps early esophageal cancer (EC) detection a global health challenge. Although noncontrast computed tomography (NC CT) is widely accessible, the esophagus is a hollow tubular structure prone to collapse and motion artifacts, making small early malignant lesions difficult to distinguish from normal tissue. Here we developed the Esophageal AI-Guided malignant Lesion Evaluation (EAGLE) model to detect precancerous lesions and cancer from chest NC CT, a task historically considered impossible. EAGLE was trained on 6,813 patients from two centers and validated across 12 centers in three countries involving 80,612 patients in opportunistic and population-based screening settings. For opportunistic screening on existing CT scans, multicenter external test cohorts (eight centers, n = 11,466) achieved 98.5% specificity, with 90.0% sensitivity for cancer and 52.5% for precancerous lesions; low-dose CT (LDCT) validation (two centers, n = 1,607) showed comparable performance, supporting EC screening through lung-cancer screening programs. Calibration in a real-world cohort (three centers, n = 35,402) reduced false positives by 72.7% while preserving sensitivity; prospective hospital validation (n = 17,446) achieved a 42.2% PPV, and real-world low-dose screening (n = 10,959) reached 99.94% specificity. EAGLE also detected precancerous lesions—in paired CT–endoscopy cohorts (two centers, n = 702), sensitivities were 65.0% for precancerous lesions and 78.4% for stage I EC at a higher-sensitivity operating point. Exploratory analyses of a prospectively enrolled cohort suggest that referring high-risk individuals for endoscopy could improve screening efficiency. In conclusion, EAGLE has the potential to serve as a scalable tool for early EC screening. Chictr.org.cn identifier: ChiCTR2300074806.

  • CFINet: Cross-Modality MRI Feature Interaction Network for Pseudoprogression Prediction of Glioblastoma

    by Ya Lv, Jin Liu, Xu Tian, Pei Yang, Yi Pan

    Pseudoprogression (PSP) is a related reaction of glioblastoma treatment, and misdiagnosis can lead to unnecessary intervention. Magnetic resonance imaging (MRI) provides cross-modality images for PSP prediction studies. However, how to effectively use the complementary information between the cross-modality MRI to improve PSP prediction is still a challenging task. To address this challenge, we propose a cross-modality feature interaction network for PSP prediction. Firstly, we propose a triple-branch multi-scale module to extract low-order feature representations and a skip-connection multi-scale module to extract high-order feature representations. Then, a cross-modality interaction module based on attention mechanism is designed to make the complementary information between cross-modality MRI fully interact. Finally, the high-order cross-modality interaction information is fed into a multi-layer perceptron to achieve the PSP prediction task. We evaluate the proposed network on a private dataset with 52 subjects from Hunan Cancer Hospital and validate it on a private dataset with 30 subjects from Xiangya Hospital. The accuracy of our proposed network on the datasets is 0.954 and 0.929, respectively, which is better than most typical convolutional neural network and interaction methods.


  • Open Access

    Real-world applications of deep convolutional neural networks in diagnostic cancer imaging

    by Hesham Elhalawani, Pei Yang, Mohamed Abazeed, Chirag Shah, Abdallah S. R. Mohamed, Charles R. Thomas Jr, Clifton D. Fuller, Reid F. Thompson

    Real-world applications of deep convolutional neural networks in diagnostic cancer imaging


  • Open Access

    Radiation-induced lymphopenia correlates with survival in nasopharyngeal carcinoma: impact of treatment modality and the baseline lymphocyte count

    by Xiaoxue Xie, Shenglan Gong, Hekun Jin, Pei Yang, Ting Xu, Yilong Cai, Chengxian Guo, Rong Zhang, Fan Lou, Wenjuan Yang, et al.

    Background and purpose: We evaluated the relationship between patient-, tumor-, and treatment-related features and radiation-induced lymphopenia (RIL) and evaluated the correlation between RIL and survival outcome in NPC patients to help improve the treatment strategy.Methods: This retrospective study included 374 patients with stage II-IVa NPC who had been treated with definitive RT and were enrolled from 2004 to 2015; The associations between the G3–4 RIL (absolute lymphocyte count, ALC < 0.5 × 109cells/L) during RT and patient-, tumor-, and treatment-related factors were assessed using Cox regression analyses. The correlation between ALC nadir and survival was examined using a Kaplan-Meier analysis, compared with the log-rank test, and confirmed by a Cox proportional hazards analysis.Results: In the multivariate analysis, lower baseline ALC and intensity modulated radiation therapy (IMRT) (vs. 2 dimensional-conformal radiation therapy,2D-CRT) were identified as 2 independent factors that were associated with G3–4 RIL. In the multivariate survival analysis, patients with G3–4 ALC nadir had longer local recurrence-free survival durations (LRFS) (vs. G0–2 nadir, HR = 0.548,P = 0.005) and longer progression-free survival durations (PFS) (vs. G0–2 nadir, HR = 0.676,P = 0.022), while patients with G4 ALC nadir had a shorter distant-metastasis-free survival duration (DMFS) (vs. G0–2 nadir, hazard ratio [HR] = 2.567,P = 0.037).Conclusions: In the study, lymphopenia during RT were affected by baseline ALC and RT modality independently. Moreover, G3–4 ALC nadir was independently linked with longer PFS and LRFS durations, while G4 ALC nadir was independently linked with a shorter DMFS duration.


  • Open Access

    Predicting pathological complete response by comparing MRI‐based radiomics pre‐ and postneoadjuvant radiotherapy for locally advanced rectal cancer

    by Yuqiang Li, Wenxue Liu, Qian Pei, Lilan Zhao, Cenap Güngör, Hong Zhu, Xiangping Song, Chenglong Li, Zhongyi Zhou, Yang Xu, et al.

    Background: Total mesorectal excision following neoadjuvant chemoradiotherapy (nCRT) is recommended in the latest treatment of locally advanced rectal cancer (LARC).Objective: To predict whether patients with LARC can achieve pathologic complete response (pCR), comparing MRI‐based radiomics between before and after neoadjuvant radiotherapy (nRT) was performed.Methods: One hundred and sixty‐five MRI‐based radiomics features in axial T2‐weighted images were obtained quantitatively from Imaging Biomarker Explorer Software. The specific features of conventional and developing radiomics were selected with the analysis of least absolute shrinkage and selection operator logistic regression, of which the predictive performance was analyzed with receiver operating curve and calibration curve, and applied to an independent cohort.Results: One hundred and thirty‐one target patients were enrolled in the present study. A radiomics signature founded on seven radiomics features was generated in the primary cohort. A remarkable difference about Rad‐score between pCR and non‐pCR group occurred in both of primary (P < .001) or validation cohorts (P < .001). The value of area under the curves was 0.92 (95% CI, 0.86‐0.99) and 0.87 (95% CI, 0.74‐1.00) in the primary and validation cohorts, respectively. The Rad‐score (OR = 23.581; P < .001) from multivariate logistic regression analysis was significant as an independent factor of pCR.Conclusion: Our predictive model based on radiomics features was an independent predictor for pCR in LARC and could be a candidate in clinical practice.


  • Open Access

    Machine Learning Applications in Head and Neck Radiation Oncology: Lessons From Open-Source Radiomics Challenges

    by Hesham Elhalawani, Timothy A Lin, Stefania Volpe, Abdallah S R Mohamed, Aubrey L White, James Zafereo, Andrew J Wong, Joel E Berends, Shady AboHashem, Bowman Williams, et al.

    Radiomics leverages existing image datasets to provide non-visible data extraction via image post-processing, with the aim of identifying prognostic, and predictive imaging features at a sub-region of interest level. However, the application of radiomics is hampered by several challenges such as lack of image acquisition/analysis method standardization, impeding generalizability. As of yet, radiomics remains intriguing, but not clinically validated. We aimed to test the feasibility of a non-custom-constructed platform for disseminating existing large, standardized databases across institutions for promoting radiomics studies. Hence, University of Texas MD Anderson Cancer Center organized two public radiomics challenges in head and neck radiation oncology domain. This was done in conjunction with MICCAI 2016 satellite symposium using Kaggle-in-Class, a machine-learning and predictive analytics platform. We drew on clinical data matched to radiomics data derived from diagnostic contrast-enhanced computed tomography (CECT) images in a dataset of 315 patients with oropharyngeal cancer. Contestants were tasked to develop models for (i) classifying patients according to their human papillomavirus status, or (ii) predicting local tumor recurrence, following radiotherapy. Data were split into training, and test sets. Seventeen teams from various professional domains participated in one or both of the challenges. This review paper was based on the contestants' feedback; provided by 8 contestants only (47%). Six contestants (75%) incorporated extracted radiomics features into their predictive model building, either alone (n = 5; 62.5%), as was the case with the winner of the “HPV” challenge, or in conjunction with matched clinical attributes (n = 2; 25%). Only 23% of contestants, notably, including the winner of the “local recurrence” challenge, built their model relying solely on clinical data. In addition to the value of the integration of machine learning into clinical decision-making, our experience sheds light on challenges in sharing and directing existing datasets toward clinical applications of radiomics, including hyper-dimensionality of the clinical/imaging data attributes. Our experience may help guide researchers to create a framework for sharing and reuse of already published data that we believe will ultimately accelerate the pace of clinical applications of radiomics; both in challenge or clinical settings.


  • Log odds of positive lymph nodes may predict survival benefit in patients with node-positive non-small cell lung cancer

    by Weiye Deng, Ting Xu, Yifan Wang, Yujin Xu, Pei Yang, Daniel Gomez, Zhongxing Liao