Truace
Subscribe
The Index recomputed live from the record

What the evidence says.What the public feels.

The record holds 935 sourced gains and 771 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.

935 gains
771 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.

1,706 results
Show filters and sorting

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

AI gains · 935

393
HealthStableModerate evidence · 1 source

AI systems can generate rapid visual summaries directly from musculoskeletal radiology report text, with the best model producing clinically useful images in a majority of tested cases.

On September 4, 2026, a peer-reviewed study in Academic Radiology evaluated 25 musculoskeletal imaging cases where three AI systems generated visual summaries from report text alone. Two fellowship-trained MSK radiologists rated each image for anatomical accuracy and clinical usefulness, finding Gemini 3.0 Pro most consistent at 40-42% accurate and 60-65% useful, while ChatGPT and Perplexity frequently produced plausible but inaccurate images.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
93

Updated Sep 6, 2026 · TRV-2026-0996

68Index score
394
HealthStableModerate evidence · 1 source

Digital pathology and AI tools using high-resolution whole-slide images are expanding diagnostic capacity for liver cancer, liver disease, and transplantation, with growing clinical access that may help address laboratory challenges.

A September 2026 review in The Lancet Digital Health summarizes how digital pathology, image analysis, and AI, including deep learning on high-resolution whole-slide images, are being applied to liver disease, liver cancer diagnosis, and transplantation assessment. The authors describe expanding use from long-standing research applications to increasing clinical practice access.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
93

Updated Sep 6, 2026 · TRV-2026-0995

68Index score
395
HealthStableModerate evidence · 1 source

Elastic Net model distinguished three depressive symptom trajectories in older adults with chronic conditions and was deployed as an interactive web-based risk calculator for individualized risk profiling.

Using longitudinal data from 5492 older Chinese adults with chronic conditions in CHARLS, researchers identified three depressive symptom trajectories and compared 10 machine learning models, selecting 10 core predictors via bootstrap RFE and evaluating with stratified split and SHAP interpretability.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
93

Updated Sep 6, 2026 · TRV-2026-0994

68Index score
396
HealthStableModerate evidence · 1 source

Interpretable machine learning combining admission D-dimer and total bleeding volume improved prediction of 12-month functional outcome after aneurysmal subarachnoid hemorrhage, with XGBoost achieving AUC 0.904 and combined D-dimer+TBV+Hunt-Hess outperforming single markers, to inform early risk stratification.

Researchers analyzed 473 patients with aneurysmal subarachnoid hemorrhage from the retrospective PROSAH-MPC cohort to test whether admission D-dimer levels and total bleeding volume together predict long-term function. They stratified patients by D-dimer quartiles, ran multivariable logistic regression for 12-month modified Rankin Scale outcomes, selected features with Boruta, and built seven machine learning models interpreted with SHAP.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
93

Updated Sep 6, 2026 · TRV-2026-0993

68Index score

AI problems · 771

393
HealthStableModerate evidence · 1 source

The original 2019 PROBAST tool had become outdated given rapid progress in prediction modelling methodology and AI including machine learning.

Published March 24 2025 in the BMJ, this methods article describes PROBAST+AI, an updated assessment tool for prediction models built with regression or artificial intelligence methods. It splits assessment into model development and model evaluation, each organized around participants and data sources, predictors, outcome, and analysis domains.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
84

Updated Jul 24, 2026 · TRV-2026-0537

67Index score
394
PolicyStableModerate evidence · 1 source

Current automated and human moderation systems on Roblox fail to prevent young users' exposure to inappropriate content, cyberbullying, and predatory behavior.

Published 25 April 2025, this peer-reviewed case study examines Roblox as a child-focused Metaverse platform, analyzing why automated and human moderation struggles with real-time interactions and massive volumes of user-generated content and documenting failures that left young users exposed to inappropriate content and predatory risks.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
84

Updated Jul 24, 2026 · TRV-2026-0536

67Index score
395
HealthStableModerate evidence · 1 source

ML-enhanced point-of-care testing faces regulatory hurdles, reliability questions, and privacy concerns that limit widespread clinical adoption.

Published April 2, 2025 in Nature Communications, this Perspective examines how machine learning is being integrated into decentralized point-of-care testing platforms, including lateral flow, vertical flow, nucleic acid amplification, and imaging-based sensors, following a pandemic-driven shift away from centralized labs.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
84

Updated Jul 24, 2026 · TRV-2026-0532

67Index score
396
HealthStableModerate evidence · 1 source

Medical LLMs still face numerous challenges in practical applications, including hallucination, limited interpretability, and ethical concerns that hinder widespread use.

This review describes the rapid development of large language models such as GPT-4 and their growing use in medicine. By May 2025, the authors state that LLMs have been gradually implemented in clinical practice, medical research, and medical education, while still facing challenges of hallucination, interpretability, and ethics.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
84

Updated Jul 24, 2026 · TRV-2026-0526

67Index score

Recomputed live from the record · Oct 13, 2026, 3:28 AM