Browsing by Subject "Prediction model"

Now showing 1 - 5 of 5
  • Thumbnail Image
    Item
    A clinical nomogram for predicting tumor regression grade in esophageal squamous-cell carcinoma treated with immune neoadjuvant immunotherapy
    (AME Publishing Company, 2022) Yu, Yongkui; Wang, Wei; Qin, Zimin; Li, Haomiao; Liu, Qi; Ma, Haibo; Sun, Haibo; Bauer, Thomas L.; Pimiento, Jose M.; Gabriel, Emmanuel; Birdas, Thomas; Li, Yin; Xing, Wenqun; Surgery, School of Medicine
    Background: There are various treatment options for esophageal squamous cell cancer. including surgery, peri-operative chemotherapy, and radiation. More recently, neoadjuvant immunotherapy has also been shown improve outcomes. In this study, we addressed the question, "Can we predict which patients with esophageal squamous cell cancer will benefit from neoadjuvant immunotherapy?". Methods: All patients with thoracic esophageal squamous-cell carcinoma (T2N+M0-T3-4N0/+M0) (according to the eighth edition of the National Comprehensive Cancer Network guidelines) who underwent immune neoadjuvant immunochemotherapy with programmed cell death protein 1 (PD-1) combined with paclitaxel plus cisplatin or nedaplatin in the Affiliated Cancer Hospital of Zhengzhou University, China, between November 2019 and August 2021 were included in this study. All patients underwent surgical resection. We developed a response [tumor regression grade (TRG)] prediction model using the least absolute shrinkage and selection operator (LASSO) regression incorporating factors associated with response. The accuracy of the prediction model was then validated. Results: We included 79 patients who underwent neoadjuvant immunotherapy combined with chemotherapy, aged 48-78 years (62.05±6.67), including 21 males and 58 females. There were five cases of immune-related pneumonia, of which three cases were diagnosed as immune-related pneumonia during the perioperative period, and one case of immune-related thyroid dysfunction changes. After LASSO regression, the factors that were independently associated with TRG were clinical T stage before neoadjuvant therapy, clinical N stage before neoadjuvant therapy, albumin level difference from before to after neoadjuvant therapy, white blood cell (WBC) count before neoadjuvant therapy, and T stage before surgery. We constructed a prediction model, plotted the nomogram, and verified its accuracy. Its Brier score was 0.13, its calibration slope was 0.98, and its C-index was 0.90 (95% CI: 0.82-0.97). Conclusions: Our prediction model can predict the likelihood of TRG in patients with esophageal squamous cell cancer after immunotherapy combined with neoadjuvant chemotherapy. Using this prediction model, we plan to conduct a subsequent neoadjuvant radiotherapy in patients with of TRG 2-3 patients with neoadjuvant radiotherapy.
  • Thumbnail Image
    Item
    Predicting conversion of brain β-amyloid positivity in amyloid-negative individuals
    (BMC, 2022-09-12) Park, Chae Jung; Seo, Younghoon; Choe, Yeong Sim; Jang, Hyemin; Lee, Hyejoo; Kim, Jun Pyo; Alzheimer’s Disease Neuroimaging Initiative; Radiology and Imaging Sciences, School of Medicine
    Background: Cortical deposition of β-amyloid (Aβ) plaque is one of the main hallmarks of Alzheimer's disease (AD). While Aβ positivity has been the main concern so far, predicting whether Aβ (-) individuals will convert to Aβ (+) has become crucial in clinical and research aspects. In this study, we aimed to develop a classifier that predicts the conversion from Aβ (-) to Aβ (+) using artificial intelligence. Methods: Data were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort regarding patients who were initially Aβ (-). We developed an artificial neural network-based classifier with baseline age, gender, APOE ε4 genotype, and global and regional standardized uptake value ratios (SUVRs) from positron emission tomography. Ten times repeated 10-fold cross-validation was performed for model measurement, and the feature importance was assessed. To validate the prediction model, we recruited subjects at the Samsung Medical Center (SMC). Results: A total of 229 participants (53 converters) from the ADNI dataset and a total of 40 subjects (10 converters) from the SMC dataset were included. The average area under the receiver operating characteristic values of three developed models are as follows: Model 1 (age, gender, APOE ε4) of 0.674, Model 2 (age, gender, APOE ε4, global SUVR) of 0.814, and Model 3 (age, gender, APOE ε4, global and regional SUVR) of 0.841. External validation result showed an AUROC of 0.900. Conclusion: We developed prediction models regarding Aβ positivity conversion. With the growing recognition of the need for earlier intervention in AD, the results of this study are expected to contribute to the screening of early treatment candidates.
  • Thumbnail Image
    Item
    Prediction Model for Brain Metastasis in Patients With Metastatic Germ‐Cell Tumors
    (Wiley, 2025) Salous, Tareq; Ashkar, Ryan; Althouse, Sandra K.; Cary, Clint; Masterson, Timothy; Hanna, Nasser H.; King, Jennifer; Einhorn, Lawrence H.; Adra, Nabil; Medicine, School of Medicine
    Background: Brain metastasis (BM) is an independent adverse prognostic factor in metastatic germ cell tumors (mGCT). We aimed to establish an effective and practical BM prediction model. Patients and methods: Between January 1990 and September 2017, 2291 patients with mGCT who were treated at Indiana University were identified. Patients were divided into two categories: BM present (N = 154) and BM absent (N = 2137). Kaplan-Meier methods were used to analyze progression free survival (PFS) and overall survival (OS). Logistic regression was used to determine a predictive model for whether BM was present. The data was separated into training and validation datasets with equal numbers of events in each. Results: The 2-year PFS and OS for patients with versus without BM: 17% versus 65% (p < 0.001) and 62% versus 91% (p < 0.001) respectively. Among the 154 patients with BM, 64 (42%) had radiation only (whole-brain radiotherapy or gamma knife), 22 (14%) had BM-surgery only, 14 (9%) had both radiation and BM-surgery. 54 patients (35%) did not receive local therapy for BM. Stepwise selection was used to determine the best model with p < 0.15 as the entry and staying criteria. The model with the largest ROC AUC was used moving forward. The model was tested in the validation dataset. A model was generated including age at diagnosis ≥ 40, choriocarcinoma predominant histology, pre-chemotherapy hCG≥ 5000, presence of pulmonary metastases size < 3, or ≥ 3 cm, and presence of bone metastasis. Patients with score of 0, 1, 2, 3, 4, 5, 6, 7, 8 points had a 0.6%, 1.4%, 3.5%, 8.2%, 18.3%, 36%, 58%, 78%, 90% probability of having BM, respectively. Conclusions: The prediction model developed in this study demonstrated discrimination capability of predicting BM occurrence in mGCT and can be used to identify high-risk patients.
  • Thumbnail Image
    Item
    RegSNPs-intron: a computational framework for predicting pathogenic impact of intronic single nucleotide variants
    (BMC, 2019-11-28) Lin, Hai; Hargreaves, Katherine A.; Li, Rudong; Reiter, Jill L.; Wang, Yue; Mort, Matthew; Cooper, David N.; Zhou, Yaoqi; Zhang, Chi; Eadon, Michael T.; Dolan, M. Eileen; Ipe, Joseph; Skaar, Todd C.; Liu, Yunlong; Medical and Molecular Genetics, School of Medicine
    Single nucleotide variants (SNVs) in intronic regions have yet to be systematically investigated for their disease-causing potential. Using known pathogenic and neutral intronic SNVs (iSNVs) as training data, we develop the RegSNPs-intron algorithm based on a random forest classifier that integrates RNA splicing, protein structure, and evolutionary conservation features. RegSNPs-intron showed excellent performance in evaluating the pathogenic impacts of iSNVs. Using a high-throughput functional reporter assay called ASSET-seq (ASsay for Splicing using ExonTrap and sequencing), we evaluate the impact of RegSNPs-intron predictions on splicing outcome. Together, RegSNPs-intron and ASSET-seq enable effective prioritization of iSNVs for disease pathogenesis.
  • Thumbnail Image
    Item
    The pneumonia severity index: Assessment and comparison to popular machine learning classifiers
    (Elsevier, 2022) Wang , Dawei; Willis, Deanna R.; Yih, Yuehwern; Medicine, School of Medicine
    Introduction: Pneumonia is the top communicable cause of death worldwide. Accurate prognostication of patient severity with Community Acquired Pneumonia (CAP) allows better patient care and hospital management. The Pneumonia Severity Index (PSI) was developed in 1997 as a tool to guide clinical practice by stratifying the severity of patients with CAP. While the PSI has been evaluated against other clinical stratification tools, it has not been evaluated against multiple classic machine learning classifiers in various metrics over large sample size. Methods: In this paper, we evaluated and compared the prediction performance of nine classic machine learning classifiers with PSI over 34,720 adult (age 18+) patient records collected from 749 hospitals from 2009 to 2018 in the United States on Receiver Operating Characteristic (ROC) Area Under the Curve (AUC) and Average Precision (Precision-Recall AUC). Results: Machine learning classifiers, such as Random Forest, provided a statistically highly(p < 0.001) significant improvement (∼33% in PR AUC and ∼6% in ROC AUC) compared to PSI and required only 7 input values (compared to 20 parameters used in PSI). Discussion: Because of its ease of use, PSI remains a very strong clinical decision tool, but machine learning classifiers can provide better prediction accuracy performance. Comparing prediction performance across multiple metrics such as PR AUC, instead of ROC AUC alone can provide additional insight.