(Press-News.org)
Submucosal tumors (SMTs) are usually found in the stomach and esophagus during an upper endoscopy. Submucosal tunneling endoscopic resection (STER) and non-tunneling endoscopic resection (NTER) are the two most commonly used techniques in the treatment of gastric and esophageal SMTs. As novel technologies continue to shape the medical landscape, machine learning (ML) algorithms find increased application, demonstrating enhanced performance in various fields. Although some studies have evaluated the incremental value of flexible ML methods, comparisons with traditional logistic regression (LR) models are lacking.
To this end, a recent study by a team of researchers from China published in the KeAi journal Gastroenterology & Endoscopy, compared traditional regression models and ML algorithms to predict which technique performs better in surgery for submucosal tumors of the cardia.
Using key baseline predictive factors, ML algorithms and LR were conducted in 246 patients. For the ML algorithms, gradient-boosting machines, artificial neural networks, random forests, and support vector machines, were included. For small sample-sized data, a technique for k-fold cross-validation was exploited to avoid over-fitting. Meanwhile, the researchers tuned the parameters through several replications. Consequently, they quantified the discrimination (area under the curve, AUC) and predictive ability (Brier score, F1 score, specificity, sensitivity, and accuracy) of models.
“Four experts who have broad experience in STER and NTER in the upper GI tract (>1,000 cases) were asked to decide on the surgical technique for each patient. Predictors include mucosal status, growth pattern, maximum diameter, layer of origin, location, and morphology. Missing data were filled by Multiple Imputations by Chain Equations (MICE),” explained Quan-Lin Li, corresponding of the study.
The team found that LR outperformed among all groups (Brier score = 0.1398, F1 score = 0.7391, AUC = 0.8729, and predictive accuracy = 80.65 %). Morphology ranked in the top tier of all importance score lists, being the highest contributor to prediction accuracy. The direction of the gastroscope was also a key factor in most models. The other seven variables showed varying importance across different models.
“A limitation of our study is that the predictor used is relatively small, which potentially limited the performance of ML algorithms. Predictors with a higher correlation should be explored to improve ML algorithms. Besides, external validation is essential before applying prediction algorithms in clinical practice, and our study did not include external validation cohorts because of the difficulty in generalizing inconsistent clinical settings from other centers,” noted Li.
“The traditional regression approach outperformed ML algorithms for the prediction of the best surgical method in patients with SMTs. Further research is needed to validate and generalize our findings,” concluded Li.
###
Contact the author: Quan-Lin Li, Endoscopy Center and Endoscopy Research Institute, Zhongshan Hospital, Fudan University, Shanghai, China, li.quanlin@zs-hospital.sh.cn.
The publisher KeAi was established by Elsevier and China Science Publishing & Media Ltd to unfold quality research globally. In 2013, our focus shifted to open access publishing. We now proudly publish more than 100 world-class, open access, English language journals, spanning all scientific disciplines. Many of these are titles we publish in partnership with prestigious societies and academic institutions, such as the National Natural Science Foundation of China (NSFC).
END
Federated learning (FL) has emerged as a popular machine learning paradigm which allows multiple data owners to train models collaboratively with out sharing their raw datasets. It holds potential for a wide spectrum of nalytics applications on sensitive data. For example, federated learning has been applied on medical big data analysis such as disease prediction and diagnosis without revealing the patients’ private medical information to thirdparty services. It has also been exploited by banks and insurance ...
Researchers have newly identified a universal, essential biomarker for the childhood cancer neuroblastoma – and a potential new target for treatment.
Neuroblastoma accounts for 15% of all pediatric cancer deaths and is the most common source of childhood tumors outside of brain cancer. The disease develops in early nerve tissue, usually in and around the adrenal glands, and typically affects children under age five. High-risk cases have a five-year survival rate of just 50%.
Led by UC San Francisco, researchers suspected the oncoprotein AF1q, which is known to play a role in leukemia and solid tumor progression, might be important in tumors of neural origin ...
Scientists at UC San Francisco may have discovered a new way to test for autism by measuring how children’s eyes move when they turn their heads.
They found that kids who carry a variant of a gene that is associated with severe autism are hypersensitive to this motion.
The gene, SCN2A, makes an ion channel that is found throughout the brain, including the region that coordinates movement, called the cerebellum. Ion channels allow electrical charges in and out of cells and are fundamental to how they function. Several variants ...
Why hydrogen causes steels to become brittle and crack is the great conundrum of engineers and researchers looking to develop large-scale transport and storage solutions for the hydrogen age – an era which Australia hopes to lead by 2030.
They may now be one step closer to understanding how hydrogen affects steels, thanks to new University of Sydney research. The researchers found adding the chemical element molybdenum to steel reinforced with metal carbides markedly enhances its ability to ...
Concerted international efforts are urgently needed to meet the targets set out in the 2030 Global Alcohol Action Plan (GAAP) and avert “dire consequences” for low and middle income countries, where alcohol markets are expanding, warn health scientists in the open access journal BMJ Global Health.
A lack of progress on alcohol and health in the wake of the 2010 Global Strategy for Reducing the Harmful Use of Alcohol prompted the 75th World Health Assembly to initiate the Global Alcohol Action Plan 2022-30 and declare alcohol a public health priority, ...
The common inflammatory condition sinusitis is linked to a 40% heightened risk of a subsequent diagnosis of rheumatic disease, particularly in the 5 to 10 years preceding the start of symptoms, finds research published in the open access journal RMD Open.
The risks seem to be greatest for a blood clotting disorder (antiphospholipid syndrome) and a condition that affects the body’s production of fluids, such as spit and tears, known as Sjögren’s syndrome, the findings indicate.
Sinusitis refers to inflammation of the lining of ...
Patients whose asthma is poorly controlled have eight times excess greenhouse gas emissions compared with those whose condition is well controlled—equivalent to that produced by 124,000 homes each year in the UK—indicates the first study of its kind, published online in the journal Thorax.
Improving the care of asthma patients could achieve substantial carbon emissions savings, and help the NHS meet its net zero target, say the researchers.
Healthcare is a major contributor to greenhouse gas emissions and in 2020 the NHS set an ambitious target of reducing its carbon footprint by 80% over the next 15 years, with the aim of reaching net zero by 2045, ...
In the first study to use whole genome sequencing to examine tandem repeat expansions in heart conditions, scientists at The Hospital for Sick Children (SickKids) have laid the groundwork for early detection of and future precision therapies for cardiomyopathy.
Cardiomyopathy is an inherited heart condition that impacts up to one in 500 individuals. The condition affects the structure and function of the heart and can ultimately lead to heart failure.
The SickKids-led study, published in eBioMedicine, part of The Lancet Discovery Science, indicates that tandem repeats – a form of genetic variation – are more often expanded ...
“This [Tic Tac-shaped object that] had just traveled 60 miles in…less than a minute, was far superior in performance to my brand-new F/A-18F and did not operate with any of the known aerodynamic principles that we expect for objects that fly in our atmosphere.”
In July of 2023, retired commander in the U.S. Navy David Fravor testified to the House Oversight Committee about a mysterious, Tic Tac-shaped object that he and three others observed over the Pacific Ocean in 2004. The congressional hearings riveted ...
Interventions such as mindfulness and cognitive behavioural therapy (CBT), could be an effective treatment option for menopause-related mood symptoms, memory and concentration problems, finds a new study by UCL researchers.
The research, published in the Journal of Affective Disorders, is the most up-to-date study of its kind, providing a meta-analysis of 30 studies involving 3,501 women who were going through the menopause in 14 countries, including the UK, USA, Iran, Australia, and China.
Lead author, Professor Aimee Spector (UCL Psychology & Language Sciences), said: “Women can spend a notable number of years in their ...