Article
Biochemistry & Molecular Biology
Fatemeh Eshari, Fahime Momeni, Amirreza Faraj Nezhadi, Soudabeh Shemehsavar, Mehran Habibi-Rezaei
Summary: A novel machine-learning approach based on logistic regression (LR) is used to predict protein aggregation propensity (PAP) using a dataset of hexapeptides and eight physiochemical features. The LR model, combined with sequence and feature information, achieves high accuracy and outperforms other existing methods in PAP prediction.
INTERNATIONAL JOURNAL OF BIOLOGICAL MACROMOLECULES
(2023)
Article
Construction & Building Technology
E. U. Eyo, S. J. Abbey
Summary: The study applied various machine learning models to predict the strength of soils improved by substituting materials, with tree-based and meta-ensemble models showing higher accuracy. Multiclass elements were used for analysis across multiple cross-validation methods, and ensemble methods were found to be more versatile in regression and multiclass classification tasks.
CONSTRUCTION AND BUILDING MATERIALS
(2021)
Article
Oncology
Takeshi Murata, Masayuki Yoshida, Sho Shiino, Ayumi Ogawa, Chikashi Watase, Kaishi Satomi, Kenjiro Jimbo, Akiko Maeshima, Eriko Iwamoto, Shin Takayama, Akihiko Suto
Summary: This study evaluated the impact of progesterone receptor (PR) status on the prognosis of breast cancer after isolated locoregional recurrence (ILRR) and developed a risk prediction model based on clinicopathologic factors. The model classified patients into four groups based on the number of identified risk factors and showed significant variation in distant metastasis-free survival (DMFS) among the groups.
BREAST CANCER RESEARCH AND TREATMENT
(2023)
Article
Computer Science, Artificial Intelligence
Jinping Liu, Xiaoqiang Wu, Yongming Xie, Zhaohui Tang, Yongfang Xie, Subo Gong
Summary: Machine learning has made significant advancements in medical diagnosis and prediction, but the lack of interpretability poses challenges. This study proposes a small samples-oriented interpretable machine learning model and demonstrates its effectiveness and superiority through experiments on medical datasets and real-world applications.
EXPERT SYSTEMS WITH APPLICATIONS
(2024)
Article
Acoustics
Eugen Divjak, Gordana Ivanac, Niko Radovic, Iva Biondic Spoljar, Slavica Sovic, Valentina Bahnik, Boris Brkljacic
Summary: Second-look US using shear-wave elastography (SWE) can help differentiate between benign and malignant changes in the postoperative breast. El(max) value of 171.2 kPa shows sensitivity and specificity for detecting carcinoma recurrence. Restricted diffusion remains a significant independent predictor of recurrence.
ULTRASCHALL IN DER MEDIZIN
(2022)
Article
Computer Science, Information Systems
Manuel Jimenez-Lazaro, Juan Luis Herrera, Javier Berrocal, Jaime Galan-Jimenez
Summary: This paper proposes a machine learning solution to predict energy-efficient network configurations in SDN, without the need for optimal or heuristic solutions, resulting in significantly reduced computation time.
Article
Biochemistry & Molecular Biology
Hsiao-Yun Chao, Chin-Chieh Wu, Avichandra Singh, Andrew Shedd, Jon Wolfshohl, Eric H. Chou, Yhu-Chering Huang, Kuan-Fu Chen
Summary: This study developed and validated a prediction model for 28-day mortality in patients with infection using machine learning algorithms. The results showed that the random forest model outperformed other models and logistic regression in predicting mortality.
Article
Multidisciplinary Sciences
Vungsovanreach Kong, Oui Somakhamixay, Wan-Sup Cho, Gilwon Kang, Heesun Won, HyungChul Rah, Heui Je Bang
Summary: Acute coronary syndrome (ACS) is a significant issue in global public health, with a high recurrence risk. This study developed a machine learning model to provide personalized ACS recurrence risk probabilities for each patient. The model demonstrated good performance and identified specific factors contributing to ACS recurrence.
Article
Biology
Wojciech Ksiazek, Michal Gandor, Pawe Plawiak
Summary: This paper presents a novel diagnostics approach for hepatocellular carcinoma (HCC) using machine learning techniques, combining logistic regression with genetic algorithms. Through three experiments, the approach was shown to achieve a classification accuracy of 94.55% and an f1-score of 93.56%. The results indicate that machine learning techniques optimized by the proposed concept can provide a new and accurate approach in HCC diagnosis.
COMPUTERS IN BIOLOGY AND MEDICINE
(2021)
Article
Biology
Taqwa F. Shaban, Mahmoud Y. Alkawareek
Summary: Three optimized models based on machine learning were developed to predict in vitro antibiofilm activity of antibiotics, with factors such as minimum inhibitory concentration, bacterial Gram type, and biofilm formation method being significant predictors of prediction accuracy.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Allergy
Louise Cunningham, Clarisse Ganier, Felicity Ferguson, Ian R. White, Fiona M. Watt, John McFadden, Magnus D. Lynch
Summary: The study showed that the risk of a clinically relevant positive patch test can be predicted using clinical and demographic parameters. The gradient boosting method outperforms logistic regression in certain predictions, indicating the relevance of complex nonlinear interactions in risk prediction.
CONTACT DERMATITIS
(2022)
Article
Operations Research & Management Science
Erdinc Akyildirim, Ahmet Goncu, Ahmet Sensoy
Summary: This study analyzes the predictability of twelve cryptocurrencies using machine learning classification algorithms at different frequencies, showing that price trends can be predicted to a certain degree. Support vector machines demonstrate the best predictive accuracy, reaching an average accuracy of 55-65%.
ANNALS OF OPERATIONS RESEARCH
(2021)
Article
Respiratory System
V Riveiro-Blanco, C. Pou-Alvarez, L. Ferreiro, M. E. Toubes, J. Quiroga-Martinez, J. Suarez-Antelo, J. M. Garcia-Prim, J. E. Rivo-Vazquez, R. Castro-Calvo, F. J. Gonzalez-Barcala, F. Gude, L. Valdes
Summary: The study aimed to develop a risk assessment model to predict the probability of recurrence in patients with primary spontaneous pneumothorax. The presence of blebs/bullae, failure to perform chest drainage, and low levels of hemoglobin and leukocytes were associated with a higher risk of recurrence.
Article
Oncology
Samira Dehdar, Khodakaram Salimifard, Reza Mohammadi, Maryam Marzban, Sara Saadatmand, Mohammad Fararouei, Mostafa Dianati-Nasab
Summary: This study aimed to identify the predicting factors for delayed breast cancer diagnosis in women in Iran. Machine learning methods and statistical analyses were used to analyze the data of 630 women with confirmed breast cancer. The study found that urban residency, breast disease history, and other comorbidities were the main predicting factors for delayed breast cancer diagnosis.
FRONTIERS IN ONCOLOGY
(2023)
Article
Chemistry, Physical
Yi-Fan Zhang, Wei Ren, Wei-Li Wang, Nan Li, Yu-Xin Zhang, Xue-Mei Li, Wen -Hui Li
Summary: With the development of artificial intelligence, machine learning has found various applications in the field of materials. The scarcity of data on high-entropy alloy mechanical properties makes it challenging to balance generalizability and interpretability in data-driven predictive models. A machine learning model was developed based on hardness data of the Al-Co-Cr-Cu-Fe-Ni system, and model ensembling was performed using four algorithms. The model showed high accuracy in predicting HEA hardness value and could be applied to other HEA systems as well as low hardness CrFeNi alloys.
JOURNAL OF ALLOYS AND COMPOUNDS
(2023)
Article
Rheumatology
Michelle M. A. Kip, Sytze de Roock, Inge van den Berg, Gillian Currie, Deborah A. Marshall, Luiza R. Grazziotin, Marinka Twilt, Rae S. M. Yeung, Susanne M. Benseler, Sebastiaan J. Vastert, Nico Wulffraat, Joost F. Swart, Maarten J. IJzerman
Summary: This study quantifies the costs of hospital-associated care for juvenile idiopathic arthritis (JIA) and finds significant variations in costs among individuals and subtypes. Systemic JIA has the highest annual costs, primarily attributed to medication, and costs are highest in the first month after JIA diagnosis.
ARTHRITIS CARE & RESEARCH
(2022)
Article
Economics
Koen Degeling, Maarten J. IJzerman, Catharina G. M. Groothuis-Oudshoorn, Mira D. Franken, Miriam Koopman, Mark S. Clements, Hendrik Koffijberg
Summary: This study aimed to provide guidance on modeling approaches for implementing competing events in discrete event simulations based on censored individual patient data (IPD). Two modeling approaches were compared, and their performance was assessed in terms of event incidence difference and time-to-event distribution. The study found that the level of censoring, number of events, and distribution overlap impacted the performance. Differences in cost-effectiveness estimates were also observed with increasing levels of censoring. Modelers should consider data characteristics when selecting modeling approaches and validate the results appropriately.
Review
Hematology
Martin Vu, Koen Degeling, Ella R. Thompson, Piers Blombery, David Westerman, Maarten J. IJzerman
Summary: This study systematically reviewed the economic evidence for molecular biomarker tests in hematological malignancies and found that while there are promising health economic results, the research in this area is currently limited, with many applications of technological advances yet to be evaluated.
EUROPEAN JOURNAL OF HAEMATOLOGY
(2022)
Article
Multidisciplinary Sciences
Frederik A. van Delft, Milou Schuurbiers, Mirte Muller, Sjaak A. Burgers, Huub H. van Rossum, Maarten J. IJzerman, Hendrik Koffijberg, Michel M. van den Heuvel
Summary: This study compares nine prediction methods using longitudinal tumor biomarker data to accurately and with a low false positive rate predict non-response to immunotherapy in NSCLC patients. The study evaluated nine models of varying complexity and demonstrated the usefulness of longitudinal biomarker data in predicting treatment response.
Article
Health Policy & Services
Joost J. Enzing, Saskia Knies, Jop Engel, Maarten J. IJzerman, Beate Sander, Rick Vreman, Bert Boer, Werner B. F. Brouwer
Summary: Despite the high attention given to manufacturers' costs in relation to drug prices in public and academic debates, this issue does not seem to be explicitly and systematically considered in reimbursement reports for expensive drugs.
COST EFFECTIVENESS AND RESOURCE ALLOCATION
(2022)
Review
Health Policy & Services
Victoria Freeman, Suzanne Hughes, Chelsea Carle, Denise Campbell, Sam Egger, Harriet Hui, Sarsha Yap, Silvia Deandrea, Michael Caruana, Tonia C. Onyeka, Maarten J. IJzerman, Ophira Ginsburg, Freddie Bray, Richard Sullivan, Ajay Aggarwal, Stuart J. Peacock, Kelvin K. W. Chan, Timothy P. Hanna, Isabelle Soerjomataram, Dianne L. O'Connell, Julia Steinberg, Karen Canfell
Summary: This study conducted a systematic review on the risk of COVID-19-related death for patients with and without cancer. The results showed that individuals with pre-existing cancer diagnosis have a higher risk of COVID-19-related death, although the risk is reduced after adjusting for age.
JOURNAL OF CANCER POLICY
(2022)
Review
Health Policy & Services
Chelsea Carle, Suzanne Hughes, Victoria Freeman, Denise Campbell, Sam Egger, Michael Caruana, Harriet Hui, Sarsha Yap, Silvia Deandrea, Tonia C. Onyeka, Maarten J. IJzerman, Ophira Ginsburg, Freddie Bray, Richard Sullivan, Ajay Aggarwal, Stuart J. Peacock, Kelvin K. W. Chan, Timothy P. Hanna, Isabelle Soerjomataram, Dianne L. O'Connell, Karen Canfell, Julia Steinberg
Summary: The early literature on susceptibility to SARS-CoV-2/COVID-19 for people with cancer is characterized by biases and limited data. To provide high-quality evidence for decision-making, studies should control for other potential modifiers of infection risk and perform stratified analyses.
JOURNAL OF CANCER POLICY
(2022)
Article
Health Care Sciences & Services
Deon Lingervelder, Michelle M. A. Kip, Eva D. Wiese, Hendrik Koffijberg, Maarten J. Ijzerman, Ron Kusters
Summary: This study investigated the experiences of chronic care patients with blood sampling and their expectations of an at-home blood-sampling device. The results showed that the majority of patients prefer using such a device to monitor their chronic disease. Cost analysis indicated that implementing an at-home blood-sampling device increases the cost of phlebotomy, but reduces overall societal costs mainly due to limiting productivity loss.
BMC HEALTH SERVICES RESEARCH
(2022)
Review
Multidisciplinary Sciences
Yoon-Jung Kang, Sophie O'Haire, Fanny Franchini, Maarten IJzerman, John Zalcberg, Finlay Macrae, Karen Canfell, Julia Steinberg
Summary: This study provides an overview and meta-analysis of the prevalence of dMMR/MSI/high TMB in different cancers. The results show variations in the prevalence of these biomarkers across different cancer types. These findings are important for forecasting the budget impact of drug approvals based on these pan-tumour biomarkers in health technology assessments.
SCIENTIFIC REPORTS
(2022)
Article
Health Care Sciences & Services
Amanda Pereira-Salgado, Angelyn Anton, Fanny Franchini, Robert K. Mahar, Edmond M. Kwan, Shirley Wong, Julia Shapiro, Andrew Weickhardt, Arun A. Azad, Lavinia Spain, Ashray Gunjur, Javier Torres, Phillip Parente, Francis Parnis, Jeffrey Goh, Christopher Steer, Stephen Brown, Peter Gibbs, Ben Tran, Maarten IJzerman
Summary: The health economic outcomes of real-world treatment sequencing of androgen receptor-targeted agents (ARTA) and docetaxel (DOC) are unclear.
EXPERT REVIEW OF PHARMACOECONOMICS & OUTCOMES RESEARCH
(2023)
Review
Oncology
Allison Drosdowsky, Karen E. Lamb, Rebecca J. Bergin, Lucy Boyd, Kristi Milley, Maarten J. IJzerman, Jon D. Emery
Summary: Research on timely diagnosis and treatment of colorectal cancer is important for improving patient outcomes. This systematic review identified methodological issues in previous research, including arbitrary categorization and lack of adjustment for confounders. Recommendations include avoiding artificial categorization, ensuring proper sequencing of key events, and using theoretical frameworks to detect and reduce bias.
CANCER EPIDEMIOLOGY
(2023)
Article
Rheumatology
Gillian R. Currie, Catherina G. M. Groothuis-Oudshoorn, Marinka Twilt, Michelle M. A. Kip, Maarten J. IJzerman, Susanne M. Benseler, Joost F. Swart, Sebastiaan J. Vastert, Nico M. Wulffraat, Rae Yeung, Deborah A. Marshall
Summary: The care for JIA patients has changed in the biologics era, however, biologics have important risks and are expensive. Flares after biologic withdrawal are common, but there is little clinical guidance on which patients can safely discontinue their biologics. We conducted a survey to identify the characteristics that are important to pediatric rheumatologists when considering withdrawal of biologics.
CLINICAL RHEUMATOLOGY
(2023)
Article
Oncology
Edwin Cuppen, Olivier Elemento, Richard Rosenquist, Svetlana Nikic, Maarten IJzerman, Isabelle Durand Zaleski, Geert Frederix, Lars-Ake Levin, Charles G. Mullighan, Reinhard Buettner, Trevor J. Pugh, Sean Grimmond, Carlos Caldas, Fabrice Andre, Ilse Custers, Elias Campo, Hans van Snellenberg, Anna Schuh, Hidewaki Nakagawa, Christof von Kalle, Torsten Haferlach, Stefan Froehling, Vaidehi Jobanputra
Summary: The combination of whole-genome and transcriptome sequencing (WGTS) is a comprehensive precision diagnostic test that is expected to transform diagnosis and treatment for cancer patients. However, there are barriers to the implementation and widespread adoption of this test, including considerations of utility in different cancer types, cost-effectiveness and affordability.
JCO PRECISION ONCOLOGY
(2022)
Article
Economics
William Padula, Noemi Kreif, David J. Vanness, Blythe Adamson, Juan-David Rueda, Federico Felizzi, Pall Jonsson, Maarten J. IJzerman, Atul Butte, William Crown
Summary: Advances in machine learning and artificial intelligence have great potential benefits in healthcare, including health economics and outcomes research. Machine learning can enhance these areas by analyzing big data effectively. However, a lack of transparency in how machine learning methods deliver solutions, especially in unsupervised circumstances, increases risk when using machine learning results.
Article
Health Care Sciences & Services
Gillian R. Currie, Tram Pham, Marinka Twilt, Maarten J. IJzerman, Pauline M. Hull, Michelle M. A. Kip, Susanne M. Benseler, Glen S. Hazlewood, Rae S. M. Yeung, Nico M. Wulffraat, Joost F. Swart, Sebastian J. Vastert, Deborah A. Marshall
Summary: This study explores pediatric rheumatologists' approaches to treatment decision making for biologics in JIA, identifying attributes influencing initiation and tapering. Five pediatric rheumatologists participated in interviews, outlining varied tapering strategies. Fourteen attributes were selected for a BWS survey, covering patient characteristics and contextual factors. Additional research is needed to align these characteristics with patient and parent preferences.
PATIENT-PATIENT CENTERED OUTCOMES RESEARCH
(2022)