Truace
Subscribe
The Index recomputed live from the record

What the evidence says.What the public feels.

The record holds 932 sourced gains and 770 sourced problems, averaging 68 and 66 on the index score. Readers have logged 22 public signals on the Pulse, which is kept apart and never counted as evidence.

932 gains
770 problems
Every sourced claim in the Index, one square each, shaded by the strength of its evidence. High Moderate Emerging

Ranks distinct AI gain and problem claims from the published record. Scores reward impact, independent source strength, scale, confidence, and recency.

1,702 results
Show filters and sorting

Download every matching row, not just this page:Export CSVExport JSON

AI gains · 932

89
HealthNewly addedModerate evidence · 1 source

Retrospective validation of Harrison Enterprise CTB on 3424 ED scans showed 98.3% NPV for urgent findings, with 64% of encounters identified as true negatives potentially eligible for expedited disposition and no remaining false negatives requiring urgent intervention during the index presentation.

Researchers retrospectively tested an AI model that classifies non-contrast CT brain scans as urgent or non-urgent on 3424 consecutive adult scans from a quaternary emergency department in 2024. Against consultant radiologist reports as reference, the model achieved 85.4% sensitivity, 69.1% specificity and 98.3% NPV, with 64% of encounters identified as true negatives potentially eligible for expedited disposition.

Impact 30%
69
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
98

Updated Oct 1, 2026 · TRV-2026-1238

74Index score
90
HealthStableModerate evidence · 1 source

PPV for high-confidence was superior to low-confidence detections (99.1% vs 59.1%, p Conclusion The algorithm showed good diagnostic performance for extremity fracture detection on radiographs.

Purpose To estimate diagnostic performance of a deep learning algorithm for extremity fracture detection on radiographs in patients aged ≥ 2 years using a refined reference standard. Secondary, to compare positive predictive value (PPV) by the algorithm's built-in confidence level (high vs. low) and diagnostic performance between adults (≥ 18 years) and children.

Impact 30%
69
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
97

Updated Sep 23, 2026 · TRV-2026-1176

74Index score
91
HealthStableModerate evidence · 1 source

Serum Protein Glycopatterns as Biomarkers for Machine Learning Diagnosis of Major Depressive Disorder: These features were used to train seven machine-learning models, among which K-nearest neighbours (KNN) performed best, achieving 95.3% ± 2.7% accuracy and an AUC (Micro) of 0.990 ± 0.010 in a nested 5-fold cross-validation framework.

The diagnosis of major depressive disorder (MDD) currently relies on subjective clinical assessment, underscoring the need for objective biomarkers. Glycosylation, a common post-translational modification, is involved in neuroinflammation and immune regulation, both implicated in MDD pathogenesis.

Impact 30%
69
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
96

Updated Sep 22, 2026 · TRV-2026-1164

74Index score
92
HealthStableModerate evidence · 1 source

Using 1-month EHR data, the LLM achieved an accuracy of 87.0%, AUC of 0.921, F1 score of 0.579, sensitivity of 89.1%, specificity of 86.8%, and detection rate for early kidney injury of 42.2%.

Chronic kidney disease (CKD) presents a growing public health challenge in China, exacerbated by low patient awareness and limited nephrology resources. This study evaluated the potential of large language models (LLMs) to support CKD diagnosis in primary care using time-series electronic health record (EHR) data.

Impact 30%
69
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
96

Updated Sep 20, 2026 · TRV-2026-1152

74Index score

AI problems · 770

89
HealthStableModerate evidence · 1 source

ObjectiveTo critically review the current pharmacologic treatment landscape for epidermal growth factor receptor (EGFR)-mutant non-small cell lung cancer (NSCLC), focusing on clinical pharmacology, therapeutic evolution, and emerging challenges of EGFR-targeted therapies.Data SourcesPeer-reviewed literature in PubMed, Web of Science, and Embase databases (January 2000 to June 2026) was searched using keywords including "EGFR-mutant NSCLC," "EGFR-TKIs," "drug resistance," "artificial intelligence," and "precision oncology." Reference lists and clinical trial registries were also screened.Data SummaryThis review systematically examines clinical misconceptions surrounding EGFR-targeted therapies, focusing on three cognitive biases that undermine precision implementation: overgeneralization of trial efficacy to unselected populations; oversimplified sequencing that ignores clonal selection trade-offs (second-generation TKIs reduce subsequent osimertinib Progression-free survival (PFS) by ∼2 months); and inflated efficacy perceptions for Ex20ins inhibitors (aggregate objective response rate (ORR) masks far-loop responses of only 22%) and MET combinations (true benefit confined to high-expression subgroup).

ObjectiveTo critically review the current pharmacologic treatment landscape for epidermal growth factor receptor (EGFR)-mutant non-small cell lung cancer (NSCLC), focusing on clinical pharmacology, therapeutic evolution, and emerging challenges of EGFR-targeted therapies.Data SourcesPeer-reviewed literature in PubMed, Web of Science, and Embase databases (January 2000 to June 2026) was searched using keywords including "EGFR-mutant NSCLC," "EGFR-TKIs," "drug resistance," "artificial intelligence," and "precision oncology." Reference lists and clinical trial registries were also screened.Data SummaryThis review systematically examines clinical misconceptions surrounding EGFR-targeted therapies, focusing on three cognitive biases that undermine precision implementation: overgeneralization of trial efficacy to unselected populations; oversimplified sequencing that ignores clonal selection trade-offs (second-generation TKIs reduce subsequent osimertinib Progression-free survival (PFS) by ∼2 months); and inflated efficacy perceptions for Ex20ins inhibitors (aggregate objective response rate (ORR) masks far-loop responses of only 22%) and MET combinations (true benefit confined to high-expression subgroup). We also critically evaluate artificial intelligence's translational potential in molecular subtyping, resistance prediction, and drug discovery, alongside the data, modeling, and validation barriers limiting its clinical deployment.ConclusionsDespite the expanding therapeutic armamentarium, clinically meaningful precision remains constrained by persistent misconceptions and translational bottlenecks.

Impact 30%
63
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
97

Updated Sep 23, 2026 · TRV-2026-1173

72Index score
90
HealthStableModerate evidence · 1 source

Artificial Intelligence in Cerebral Small Vessel Disease Imaging: A Study-Level Cross-Sectional Analysis of Validation Status and Clinical Applicability: White matter hyperintensity segmentation/quantification was the most common task (196/463, 42.3%), followed by cerebral microbleed detection/classification (82/463, 17.7%), perivascular space/lacune assessment (67/463, 14.5%), CSVD burden/risk modeling (62/463, 13.4%), and clinical outcome prediction (56/463, 12.1%).

Artificial intelligence (AI) is increasingly used in cerebral small vessel disease (CSVD) imaging, but the extent of validation and clinical applicability across the literature remains uncertain. We performed a study-level cross-sectional analysis using a frozen Web of Science Core Collection (WoSCC) cohort supplemented by PubMed and IEEE Xplore searches to assess database coverage and the stability of findings from the WoSCC cohort.

Impact 30%
63
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
96

Updated Sep 20, 2026 · TRV-2026-1149

72Index score
91
HealthStableModerate evidence · 1 source

Routine clinical recording of remission components was sparse, with only a small fraction of eligible patients having ACT, FEV1%, or FeNO data, limiting availability of EMR data for remission assessment and generalizability.

Researchers tested whether clinical remission in asthma could be measured from routine electronic medical records in Japan by analyzing 32,258 patients starting fluticasone furoate/umeclidinium/vilanterol triple therapy in the JAMDAS database from 2020 to 2024, extracting ACT scores with SQL and exacerbation-related hospitalizations and emergency transports with a large language model.

Impact 30%
63
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
95

Updated Sep 16, 2026 · TRV-2026-1108

72Index score
92
HealthStableModerate evidence · 1 source

Up to a quarter of H&E slides in French labs showed technical preparation imperfections and up to 23.8% showed suboptimal staining, creating inconsistent inputs that undermine robust AI performance from whole slide imaging.

A peer-reviewed analysis of French national external quality assessment data from 2019-2024 and a 2024 survey found frequent H&E preparation issues: up to 25.5% of slides had technical imperfections such as sectioning, thickness, and stretching problems, and 8.3% to 23.8% had suboptimal staining with poor nucleo-cytoplasmic contrast or intensity fluctuations.

Impact 30%
63
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
95

Updated Sep 14, 2026 · TRV-2026-1087

72Index score

Recomputed live from the record · Oct 11, 2026, 9:57 AM