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

381
HealthStableModerate evidence · 1 source

Dynamic machine learning models using vital signs, lab trends and unstructured data can detect sepsis-related deterioration hours before clinical decompensation in critically ill cancer patients, supporting a shift from reactive to predictive onco-critical care.

This peer-reviewed review examines how artificial intelligence could change onco-critical care, focusing on early sepsis detection and dynamic mortality prediction for critically ill cancer patients whose complex physiology limits traditional scores like APACHE II and SOFA.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
94

Updated Sep 9, 2026 · TRV-2026-1023

68Index score
382
HealthStableModerate evidence · 1 source

Automation and AI applied to ventilator management support waveform analysis, asynchrony detection, and weaning-readiness prediction to enable more individualized lung-protective care within predefined safety limits.

Published September 9, 2026, this peer-reviewed review outlines a pragmatic framework for physiology-guided mechanical ventilation that combines advanced monitoring, proportional assist modes, and bounded automation. It notes that automation and artificial intelligence are increasingly used for waveform analysis, asynchrony detection, and weaning prediction within predefined safety limits.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
94

Updated Sep 9, 2026 · TRV-2026-1022

68Index score
383
HealthStableModerate evidence · 1 source

Three large language models achieved higher accuracy than a panel of hip preservation experts on a 21-item consensus-based questionnaire covering femoroacetabular impingement syndrome, hip dysplasia and microinstability.

A peer-reviewed study published September 7, 2026 compared three large language models to 10 international hip preservation experts on a 21-item questionnaire derived from consensus guidelines for femoroacetabular impingement, dysplasia and microinstability. Gemini scored 100%, ChatGPT 98.4% and Claude 96.8% versus 90.5% for experts, with models also showing higher intra-item agreement.

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

Updated Sep 8, 2026 · TRV-2026-1018

68Index score
384
HealthStableModerate evidence · 1 source

NDIB-Sim achieved high-fidelity long-term brain activity prediction from short-term observations with functional connectivity similarity above 0.97 and enabled subject-specific parameters that distinguish Alzheimer's patients from cognitively normal individuals.

Researchers introduced NDIB-Sim, a multimodal bidirectional physics-informed neural network that combines a data-driven module with a multiscale neural dynamics mechanism module. By publication on 2026-09-07, extensive experiments showed it could predict long-term brain activity from short-term observations with average functional connectivity similarity above 0.97 and infer individual-specific effective connectivity.

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

Updated Sep 8, 2026 · TRV-2026-1016

68Index score

AI problems · 771

381
HealthStableModerate evidence · 1 source

Deployment of multimodal machine learning for AML transfusion management is limited by data privacy protection, data standardisation across platforms, and model interpretability for clinical adoption.

This peer-reviewed review published July 24, 2026 synthesized recent progress on machine learning models that integrate multimodal big data such as electronic health records, genomic and proteomic data to guide transfusion support for acute myeloid leukaemia, a highly heterogeneous malignancy where transfusion is essential.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
85

Updated Jul 27, 2026 · TRV-2026-0577

67Index score
382
HealthStableModerate evidence · 1 source

SPECT imaging features added limited prognostic value and implementing models on real-world data did not significantly close the gap between prognostic modeling and clinical implementation.

Researchers developed and validated Random Forest and Gradient Boosting models to predict Hoehn and Yahr scores 5 years after 123I-ioflupane SPECT imaging, using harmonized data from 343 real-world patients and 134 PPMI patients with 83 overlapping features. Models using 2 years of clinical follow-up achieved the highest accuracy, driven by early H&Y scores, gait severity, and select imaging features.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
85

Updated Jul 27, 2026 · TRV-2026-0574

67Index score
383
EducationStableModerate evidence · 1 source

Integrating AI into education creates challenges that require comprehensive educator training and curriculum adaptation to align with societal structures.

A peer-reviewed discussion published February 2024 examines AI integration in education, arguing that personalized learning and support for diverse requirements including special needs students depends on developing AI literacy, prompt engineering proficiency, and critical thinking, while requiring educator training and curriculum adaptation.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
85

Updated Jul 26, 2026 · TRV-2026-0573

67Index score
384
ScienceStableModerate evidence · 1 source

Without careful dissection of how biases are encoded into AI health technologies, underlying health dataset limitations risk perpetuating existing health inequalities at scale.

On Dec 18, 2024, The Lancet Digital Health published the STANDING Together consensus recommendations, developed through a systematic review, stakeholder survey, Delphi process with 194 voters from 25 countries, public consultation, and international interviews involving over 350 representatives from 58 countries. The process produced 29 recommendations in two parts covering documentation of health datasets and use of health datasets.

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

Updated Jul 24, 2026 · TRV-2026-0551

67Index score

Recomputed live from the record · Oct 13, 2026, 1:46 AM