AI Science Tracker
EN

University of Electronic Science and Technology of China

5 discoveries from 2021 to 2025

Discoveries

A Fully Automated Multistage Deep Learning System for Lenke Classification: Enhanced Diagnostic Precision in Adolescent Idiopathic Scoliosis.

Source check: Source checked. Not independent scientific confirmation.

Finding: A multistage deep-learning pipeline automated Lenke classification for adolescent idiopathic scoliosis from clinical imaging and was compared with specialist labels.

Technology: Deep-learning model

Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.

AI contribution
Deep-learning model enabled the stated case-specific result.
Test and result
The study reports improved agreement and precision against specialist reference labels; the reviewed evaluation does not establish prospective clinical outcome benefit.
Limitation / still unknown
The reviewed source does not establish prospective benefit, broad external validity, or replication beyond the reported study.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://doi.org/10.2106/jbjs.25.01015

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Iterative learning guides discovery of a cobalt-doped nitride battery catalyst

Source check: Source checked. Not independent scientific confirmation.

Finding: Iteratively updated regression models screened transition-metal compositions and guided synthesis of three candidate battery catalysts. Cobalt-doped molybdenum nitride, Co 0.1 Mo 0.9 N, showed the best reported performance of these candidates in laboratory lithium-carbon-dioxide and lithium-air cells. Reported low overpotentials were measured at 50 milliamperes per gram; they do not establish commercial cell performance or broad model accuracy.

Technology: Iterative supervised regression; random forest; Gaussian-process regression; XGBoost; nearest-neighbor regression

Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.

AI contribution
Regression models learned catalyst-descriptor/overpotential relationships, with measurements added across five iterations before screening 15012 compositions and selecting three for synthesis.
Test and result
Three selected catalyst compositions were synthesized and characterized; Co 0.1 Mo 0.9 N was evaluated against MoN and carbon-nanotube cathode comparators in laboratory battery tests.
Limitation / still unknown
Generalization is not established by the reported 0.106-volt mean absolute error on only three selected candidates. Practical battery durability, manufacturing and device-level benefit are unestablished; reported overpotential depends on current density and cell conditions.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://doi.org/10.1021/jacs.5c15395 · https://doi.org/10.1021/jacs.5c15395.s001

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 21 Nov 2026. Review scope: Primary publisher article-front metadata and abstract, complete official author supporting information Methods/model tables/experimental characterizations and result captions, and current PubMed notice metadata. Paywalled main article body was not read; no independent replication performed..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

A Data-Efficient Framework for the Identification of Vaginitis Based on Deep Learning

Source check: Source checked. Not independent scientific confirmation.

Finding: The primary full text's Methods and Results support the assigned case-specific AI-enabled result: Informative sampling improves vaginitis identification against clinical diagnostic labels.

Technology: Data-efficient deep-learning vaginitis classifier

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
Uses informative sample selection to improve vaginitis identification against clinical diagnostic labels.
Test and result
The full text describes its data/cohort, comparison and evaluation appropriate to the represented result; the claim is limited to this study.
Limitation / still unknown
The reported evaluation does not establish prospective benefit or performance beyond the studied setting.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC8898862/

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: other. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Predicting MCI to AD Conversation Using Integrated sMRI and rs-fMRI: Machine Learning and Graph Theory Approach.

Source check: Source checked. Not independent scientific confirmation.

Finding: RSFS feature selection paired with graph/MRI measures and machine-learning classifiers distinguished MCI converters from nonconverters and MCI converters from AD in ADNI nested-validation analyses. The reported accuracies are retrospective classification results, not prospective prediction of clinical benefit. External replication and prospective utility were not established in the reviewed material.

Technology: ADNI imaging, intra-subject MCI converter/nonconverter and inter-subject MCIc/AD classifications; RSFS feature selection and SVM nested cross-validation, compared with alternative selectors/classifiers and upsampling. Results report 84.71% and 89.80% best accuracy and associated brain features.

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
ADNI imaging, intra-subject MCI converter/nonconverter and inter-subject MCIc/AD classifications; RSFS feature selection and SVM nested cross-validation, compared with alternative selectors/classifiers and upsampling. Results report 84.71% and 89.80% best accuracy and associated brain features.
Test and result
RSFS feature selection paired with graph/MRI measures and machine-learning classifiers distinguished MCI converters from nonconverters and MCI converters from AD in ADNI nested-validation analyses. The reported accuracies are retrospective classification results, not prospective prediction of clinical benefit.
Limitation / still unknown
External replication, causal effects and prospective utility were not established in the reviewed material.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://www.frontiersin.org/journals/aging-neuroscience/articles/10.3389/fnagi.2021.688926/full

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Multisite Autism Spectrum Disorder Classification Using Convolutional Neural Network Classifier and Individual Morphological Brain Networks.

Source check: Source checked. Not independent scientific confirmation.

Finding: The paper reports cross-site classification performance on its multisite imaging dataset. The classification result does not establish a clinical diagnostic tool or generalization beyond the studied sites and labels.

Technology: A convolutional neural network classified autism-spectrum disorder using individual morphological brain networks derived from multisite MRI data.

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
A convolutional neural network classified autism-spectrum disorder using individual morphological brain networks derived from multisite MRI data.
Test and result
The paper reports cross-site classification performance on its multisite imaging dataset.
Limitation / still unknown
The classification result does not establish a clinical diagnostic tool or generalization beyond the studied sites and labels.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2020.629630/full

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: linked separately.

Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open the interactive profile for the timeline, or browse all organizations and model families.