HealthContested · G 74 / P 72
Source article: A machine learning-derived sarcopenia index is associated with survival and nonrelapse mortality in DLBCL
Problem
In DLBCL patients treated with first-line immunochemotherapy, CT-measured sarcopenia in the lowest tertile of muscle mass is associated with inferior overall survival driven by nonrelapse mortality and higher risk of hematologic toxicity.
Blood AdvancesGain
Machine learning-supported body composition analysis applied to CT imaging quantifies radiologic sarcopenia and enables risk stratification for survival after first-line immunochemotherapy in newly diagnosed DLBCL.
Blood AdvancesHealthPositive state · G 77 / P 72
Source article: Advancing hirschsprung disease diagnosis: a systematic review of the development and application of artificial intelligence in histopathological analysis
Problem
69% of studies showed high risk of bias from small sample sizes, patch-level data partitioning, and no external test sets, raising concerns about overfitting and data leakage.
Pediatric Surgery InternationalGain
Deep learning models for Hirschsprung disease histopathology achieved over 90% ganglion cell detection and cut diagnostic time by 50-95%, increasing accuracy and accelerating clinical decision-making.
Pediatric Surgery InternationalCrimeContested · G 69 / P 72
Source article: Mapping agricultural fragility in India through a yield gap vulnerability framework: a national-scale machine learning assessment
Problem
Despite increased modal yields for rice and wheat, regional yield gaps continue to widen and the share of high-vulnerability districts has risen, particularly in resource-stressed regions such as the Indo-Gangetic Plains, driven by socioeconomic inequality and climate variability.
Journal of Environmental ManagementGain
The integrated ML-based Yield Gap Vulnerability framework provides a data-driven decision-support tool that can support sustainable agricultural management, spatial planning and risk reduction by identifying vulnerability hotspots for region-specific measures in India.
Journal of Environmental ManagementHealthContested · G 70 / P 66
Source article: The Role of Artificial Intelligence Models in Predicting Post-Prosthetic Facial Esthetics in Edentulous Patients: Clinical and Anthropometric Comparative Study
Problem
AI simulations failed to accurately reproduce quantitative facial anthropometric changes after denture placement despite visual similarity.
European Journal of DentistryGain
AI models Gemini and FaceApp generated post-denture facial images rated as esthetically comparable to actual clinical outcomes by patients and experts.
European Journal of DentistryHealthContested · G 67 / P 70
Source article: Automated artificial intelligence performance for longitudinal pulmonary nodule matching in lung cancer screening
Problem
Automated matching failed for 16.5% of persisting findings and performance fell to 72.8% in participants with more than five nodules, with prospective validation in diverse populations still needed.
European RadiologyGain
Automated pulmonary AI matched persisting lung nodules across 3-month LDCT scans with 83.5% success, reaching 91.8% for single-nodule cases and leaving only 1.5% of persisting findings needing manual correction, indicating potential to reduce manual tracking workload.
European RadiologyClimateContested · G 69 / P 72
Source article: Machine learning for monitoring and assessment of potentially toxic elements in soils: a synthesis of spatial validation, explainability, and uncertainty
Problem
Many ML studies of soil potentially toxic elements rely on spatially naive validation, and random cross-validation often overestimates predictive performance when spatial dependence is ignored, with incomplete uncertainty reporting.
Environmental Monitoring and AssessmentGain
Machine learning can map potentially toxic element concentrations from environmental covariates and produce exceedance-probability maps aligned with regulatory thresholds for soil-contamination management.
Environmental Monitoring and AssessmentHealthContested · G 70 / P 66
Source article: Clinical phenotyping of bloodstream infections: a review of current evidence
Problem
Studies use inconsistent phenotyping methods and provide limited validation, slowing translation of AI-derived BSI subphenotypes into routine clinical practice.
Clinical Microbiology and InfectionGain
Unsupervised machine learning applied to bloodstream infections identifies reproducible clinical subphenotypes with different mortality, supporting bedside tools for rapid phenotype assignment and personalized antimicrobial therapy.
Clinical Microbiology and InfectionHealthContested · G 70 / P 67
Source article: Association between body composition and recurrence in stage II-III colon cancer: a retrospective cohort study
Problem
Patients with stage II-III colon cancer whose CT body composition showed myosteatosis via machine learning analysis had significantly lower 5-year recurrence-free survival.
Clinical Nutrition ESPENGain
Using a machine learning model to measure CT body composition at L3 identified myosteatosis as an independent predictor of recurrence in stage II-III colon cancer.
Clinical Nutrition ESPENHealthContested · G 72 / P 74
Source article: Reinventing the echocardiography workflow: from manual quantification to artificial intelligence-driven comprehensive interpretation
Problem
Clinical use of AI in echocardiography carries risk of automation bias in high-volume settings, compounded by inconsistent performance across platforms.
Journal of Cardiovascular ImagingGain
AI integration in echocardiography workflows reduces examination time and automates measurements, enabling more comprehensive data collection while reducing sonographer fatigue.
Journal of Cardiovascular ImagingHealthContested · G 68 / P 68
Source article: Naïve adaptive immune receptor repertoires in celiac disease assessed by machine learning; impact of the HLA-DQ2.5 allotype on the TCR repertoire
Problem
After accounting for HLA-DQ2.5 enrichment, machine-learning classification of celiac disease from naive TCR repertoires was abolished, and naive BCR repertoires failed to classify disease.
ImmunogeneticsGain
Naive CD4+ TCR repertoire features enabled moderate machine-learning classification of celiac disease status and high-accuracy prediction of HLA-DQ2.5 status by publication date.
ImmunogeneticsHealthContested · G 71 / P 71
Source article: International Application of Artificial Intelligence for Lesion Detection on Digital Breast Tomosynthesis: Comparing Western and Eastern Databases
Problem
AI lesion detection models trained on Western or Eastern DBT data showed reduced performance when applied to the other population due to differences in lesion types.
Academic RadiologyGain
Transfer learning enhanced AI lesion detection performance when models were adapted from one regional DBT database to another.
Academic RadiologyHealthContested · G 72 / P 70
Source article: Predicting Synchronous Liver Metastasis in Pancreatic Cancer Using CT Radiomics and Clinical Features: A Machine Learning Approach
Problem
The linear LDA model showed unstable performance and poor discrimination in lymph-node-negative and pancreatic head tumor subgroups despite overall validation AUC.
Academic RadiologyGain
Integrated machine learning models using CT radiomics and clinical predictors achieved accurate preoperative prediction of synchronous liver metastasis in pancreatic ductal adenocarcinoma in validation, intended to assist decisions on surveillance versus biopsy or neoadjuvant therapy for indeterminate subcentimeter CT-
Academic RadiologyHealthNegative state · G 65 / P 70
Source article: Attitudes, perceptions, and UTAUT-based factors influencing the acceptance of medical artificial intelligence among Chinese oncology healthcare professionals: a national cross-sectional survey
Problem
Despite high awareness, most Chinese oncology professionals have not practically integrated medical AI, with only 17.7% reporting both hearing of and using it and 67.7% hearing but never using it.
JAMIA OpenGain
Among Chinese oncology professionals, higher trust in system reliability and higher effort expectancy were associated with stronger behavioral intention to adopt medical AI.
JAMIA OpenHealthNegative state · G 66 / P 73
Source article: Deep Learning for Synthetic Postcontrast T1-Weighted MRI: A Systematic Review With Targeted Meta-Analysis of Brain Tumor Studies
Problem
Clinical translation is limited by inconsistent evaluation, substantially lower performance on pathology-specific regions, and reliance on single-institution data with few reader studies or external validation.
American Journal of RoentgenologyGain
Systematic review and meta-analysis of 15 brain tumor studies found deep learning synthesis of postcontrast T1-weighted MRI from precontrast sequences alone is technically feasible with high whole-image similarity.
American Journal of RoentgenologyHealthContested · G 70 / P 70
Source article: Robotic Ultrasound Imaging: A Comprehensive Review of Historical Evolution, Current State-of-the-Art, and Future Perspectives
Problem
Conventional ultrasound imaging's profound reliance on operator expertise restricts reproducibility and global accessibility.
Journal of Ultrasound in MedicineGain
Robotic ultrasound systems improve reproducibility and global accessibility by decoupling the operator from the patient and using 5G telesonography to project diagnostic expertise.
Journal of Ultrasound in Medicine
ace
