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

377
ScienceStableModerate evidence · 1 source

Ensemble tree-based machine learning models predicted band gap energies across multiple halide perovskite families with strong accuracy and identified B-site and X-site properties as key descriptors to guide design of new low-toxicity materials for optoelectronics.

On 2026-09-08, a study in Physical Chemistry Chemical Physics reported machine learning models that predict band gap energies across various halide perovskite types from atomic and structural properties. The work tested ensemble tree-based algorithms including random forest, gradient boosted trees, and XGBoost and examined feature importance to link descriptors to band gaps.

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

Updated Sep 9, 2026 · TRV-2026-1028

68Index score
378
EducationStableModerate evidence · 1 source

AI systems can generate, score and interpret educational assessments that inform learner progression, with design and governance determining whether cross-cutting mechanisms function as affordances.

A 2026 conceptual review in Medical Education examined how artificial intelligence used to generate, score and interpret assessments affects validity. Using Kane's four inferences, the authors mapped threats such as prompt instability and domain shift and noted that AI assessment is advancing without formal scrutiny comparable to clinical AI.

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

Updated Sep 9, 2026 · TRV-2026-1026

68Index score
379
HealthStableModerate evidence · 1 source

A Random Forest model integrating age, KPS, tumor diameter, NLR, AGR, IDH status and Ki-67 achieved AUC 0.864 training and 0.820 validation to assist preoperative assessment of high-grade glioma.

On 2026-09-08, a peer-reviewed study reported development of an integrative clinical-molecular model for glioma malignancy grading using 400 patients from a single-center retrospective cohort. Using LASSO and machine learning, the authors built a Random Forest classifier based on seven predictors including age, KPS, tumor diameter, inflammatory ratios, IDH status and Ki-67, achieving AUC 0.864 in training and 0.820 in validation.

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

Updated Sep 9, 2026 · TRV-2026-1025

68Index score
380
HealthStableModerate evidence · 1 source

AI-assisted non-contact heart rate monitoring provides accurate, safe, and efficient neonatal assessment with strong correlation to ECG and rapid signal acquisition.

A systematic review covering January 2013 to June 2025 examined non-contact and AI-assisted neonatal heart rate monitoring versus conventional ECG. It found a progressive shift toward camera-based photoplethysmography, thermal imaging, and AI-enhanced multimodal systems that showed strong correlation with ECG, rapid acquisition, and better robustness to motion and lighting.

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

Updated Sep 9, 2026 · TRV-2026-1024

68Index score

AI problems · 771

377
HealthStableModerate evidence · 1 source

Implementing AI in radiology education is constrained by high costs, rapid pace of technological change, and risks of bias, error, and data privacy violations.

This 2026 RadioGraphics review examines how artificial intelligence, especially generative models, could be applied across radiology education from curriculum planning to implementation and evaluation using Harden's 10-step framework.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
86

Updated Aug 1, 2026 · TRV-2026-0612

67Index score
378
HealthStableModerate evidence · 1 source

Pooled performance estimates were based predominantly on internal validation, with true external validation remaining sparse, limiting confidence in generalizability.

A systematic review and meta-analysis up to September 2025 synthesized 83 studies involving at least 136,840 patients to evaluate machine learning models predicting hematoma expansion, poor functional outcome, and mortality after spontaneous intracerebral hemorrhage. Pooled analyses found that models combining clinical and radiomics features achieved the highest discrimination, with C-indexes of 0.822, 0.850, and 0.860 respectively, largely from internal validation sets.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
86

Updated Aug 1, 2026 · TRV-2026-0611

67Index score
379
HealthStableModerate evidence · 1 source

Calibration assessment showed dataset shift, so absolute predicted probabilities should be interpreted cautiously and the model should not drive CSF diversion decisions alone without prospective validation.

Investigators developed and externally validated a machine learning model to stratify risk of postoperative hydrocephalus after posterior fossa tumor resection using data from 1,073 patients treated at five tertiary centers from 2013 to 2024. After screening 30 variables, they built a three-variable preoperative model using Evans index, tumor-fourth ventricle relationship, and preoperative CSF diversion status, with SVM showing AUC 0.877 in the external cohort.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
86

Updated Jul 31, 2026 · TRV-2026-0600

67Index score
380
HealthStableModerate evidence · 1 source

Rapid integration of AI/ML introduced interconnected ethical challenges around bias, accountability, privacy, and equity that affect patient safety and trust.

By July 2026, a narrative review of 127 peer-reviewed studies from 2015-2026 examined how AI and ML are being used in pharmaceutical research and healthcare. The review found the technologies enable large-scale biomedical data analysis and data-driven decision-making while simultaneously introducing interconnected ethical challenges.

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

Updated Jul 27, 2026 · TRV-2026-0578

67Index score

Recomputed live from the record · Oct 13, 2026, 1:21 AM