Clinical trial shows new technique detects 11,000 cortical lesions in routine MRI scans — Evidence Review
Published in Communications Medicine, by researchers from University at Buffalo, Genentech
Researched by
— the AI search engine for science
Table of Contents
Researchers at the University at Buffalo have developed an AI method that unveils hidden signs of multiple sclerosis (MS) in standard MRI scans, revealing thousands of previously undetectable cortical lesions. Related studies consistently highlight the clinical importance of gray matter lesions in MS and the limitations of conventional MRI, largely supporting these new findings.
- Multiple reviews and observational studies have established that cortical lesions are frequent in MS and strongly associated with disease progression and cognitive decline, but standard MRI has struggled to detect them reliably 6 7 8 9 10 13.
- Advanced MRI techniques and quantitative imaging approaches have improved detection of gray matter damage, but they require specialized protocols not always available in routine clinical settings; the new AI method leverages existing scans, potentially broadening access 2 3 4 5 13.
- Deep learning and AI-based approaches for automated lesion detection are increasingly recognized as effective tools to reduce manual error and improve diagnostic accuracy, in line with the new study's approach 5.
Study Overview and Key Findings
The detection of cortical lesions in MS has long posed a challenge for clinicians, as standard MRI typically highlights white matter damage while missing significant pathology in the brain’s cortex. This gap has hindered the ability to fully understand disease progression, particularly the aspects most associated with cognitive impairment and disability. The new study addresses this gap by applying artificial intelligence to standard MRI datasets from a large clinical trial, revealing hidden cortical lesions and providing a more comprehensive view of MS pathology.
| Property | Value |
|---|---|
| Study Year | 2026 |
| Organization | University at Buffalo, Genentech |
| Journal Name | Communications Medicine |
| Authors | Michael G. Dwyer, Niels Bergsland, Alexander Bartnik, Dejan Jakimovski, Samantha Noteboom, Menno M. Schoonheim, Martijn D. Steenwijk, Jinglan Pei, David Clayton, Robert Zivadinov |
| Population | Participants in a phase III clinical trial for MS |
| Sample Size | n=700 |
| Methods | Randomized Controlled Trial (RCT) |
| Outcome | Detection of cortical lesions in MRI scans |
| Results | Detected over 11,000 cortical lesions across the dataset. |
Literature Review: Related Studies
To situate the new findings in the broader research landscape, we searched the Consensus database, which contains over 200 million research papers. The following search queries were used to identify relevant studies:
- multiple sclerosis MRI detection techniques
- cortical lesions prevalence in MS
- hidden damage multiple sclerosis imaging
Summary Table of Key Topics and Findings
| Topic | Key Findings |
|---|---|
| How well do current MRI techniques detect cortical and gray matter lesions in MS? | - Conventional MRI often misses cortical lesions, while advanced sequences (e.g., double inversion recovery, diffusion imaging) and quantitative MRI improve detection but are not routine in clinical practice 1 2 3 4 11 13. - Deep learning and AI-enhanced techniques improve lesion detection accuracy and efficiency, addressing manual error and time constraints 5. |
| What is the clinical significance of cortical lesions in MS progression and disability? | - Cortical lesions are strongly associated with cognitive impairment and disability progression; their accumulation correlates with worse outcomes and can predict disease evolution 6 7 8 9 10. - Early cortical demyelination is frequent and linked to inflammatory activity, emphasizing the importance of early detection 6 7 8 9 10. |
| What are the limitations and future prospects of MRI and AI in MS diagnosis? | - Quantitative and advanced MRI methods provide better insight into MS mechanisms and treatment response, but integration into clinical practice is limited by technical and logistical barriers 2 3 4 12 13. - AI and deep learning approaches offer promise for automated, scalable analysis, but clinical adoption and standardization require further research 5. |
How well do current MRI techniques detect cortical and gray matter lesions in MS?
Conventional MRI, widely used in MS diagnosis and monitoring, detects white matter lesions effectively but often fails to visualize cortical and gray matter lesions, which play a crucial role in disease progression. Advanced imaging techniques such as double inversion recovery (DIR), diffusion MRI, and quantitative approaches have improved detection rates but are not yet standard in routine clinical care. The new study’s AI-based method builds on these advances by extracting hidden cortical lesion signals from standard MRI scans, potentially making improved detection more accessible.
- Conventional MRI techniques lack sensitivity for cortical lesions, necessitating the use of advanced protocols like DIR or higher field strengths 1 2 4 13.
- Quantitative MRI and diffusion imaging provide more specific markers for underlying pathology and can uncover damage not seen on routine scans 3 11 12.
- AI and deep learning algorithms can enhance lesion detection, reduce manual workload, and minimize error, addressing limitations of human interpretation 5.
- The new study’s approach aligns with these trends, enabling retrospective analysis of existing MRI datasets to uncover hidden pathology 5 13.
What is the clinical significance of cortical lesions in MS progression and disability?
Cortical lesions have emerged as significant predictors of MS severity, disease progression, and cognitive impairment. Multiple studies have demonstrated that both the number and volume of cortical lesions correlate with disability scores, cognitive deficits, and the transition to secondary progressive MS. Early and accurate detection of these lesions is therefore critical for prognosis and treatment decisions.
- Cortical lesions are common in early-stage MS and are closely associated with meningeal inflammation and neurodegeneration 6.
- Lesion accumulation over time is linked to increased disability and cognitive decline; patients with more cortical lesions tend to have worse functional outcomes 7 8 9 10.
- High-field MRI studies confirm that cortical lesion burden is a stronger predictor of cognitive and physical disability than white matter lesion volume 9.
- The new AI method provides a tool for quantifying this hidden burden, potentially improving disease monitoring and personalized care 7 8 9 10.
What are the limitations and future prospects of MRI and AI in MS diagnosis?
While advanced imaging and AI techniques offer significant promise, widespread clinical adoption faces several hurdles. Quantitative MRI and diffusion models provide valuable insights into MS mechanisms and treatment efficacy, but require standardization and validation in larger cohorts. AI-based methods, including the one described in the new study, must be further evaluated for reliability, generalizability, and clinical utility.
- Advanced MRI methods such as relaxometry, myelin imaging, and NODDI models provide more detailed assessment of tissue integrity but are not routinely used 2 3 12.
- Deep learning and AI tools improve efficiency and accuracy, yet require careful validation before integration into clinical workflows 5.
- Future prospects include automated lesion quantification, improved patient stratification, and more precise monitoring of disease progression and treatment response 3 5 12.
- The new study contributes to this evolution by demonstrating that AI can unlock previously inaccessible data from standard imaging, paving the way for retrospective and prospective analyses 5 13.
Future Research Questions
Despite recent advances, several important questions remain regarding the optimal use of AI and imaging techniques in MS diagnosis and management. Future research should address the clinical implications of hidden cortical lesions, standardization of AI tools, and the integration of these approaches into routine practice.
| Research Question | Relevance |
|---|---|
| How does AI-based detection of cortical lesions impact clinical outcomes in MS patients? | Understanding whether identifying hidden lesions leads to better prognostication or treatment decisions is critical for clinical adoption 7 8 9 10. |
| What are the long-term implications of cortical lesion burden revealed by AI on disease progression in MS? | Longitudinal studies are needed to clarify how newly detected lesions relate to cognitive decline, disability, and MS subtype transitions over time 8 9. |
| Can AI algorithms be standardized across different MRI scanners and clinical centers? | Standardization is essential for clinical translation, ensuring reliability and reproducibility across sites and imaging protocols 3 4 5. |
| How does the presence of cortical lesions influence response to disease-modifying therapies in MS? | Investigating whether cortical lesion burden affects treatment effectiveness could inform therapy selection and patient management 3 8. |
| What are the ethical and practical implications of retrospectively analyzing legacy MRI data with AI? | Evaluating privacy, consent, and data governance issues is necessary as AI tools are applied to existing clinical and research imaging archives 5. |
This article provides a comprehensive, evidence-based summary of the new AI method for detecting hidden MS lesions and situates it within the broader context of MS imaging research. The findings align with a growing emphasis on the importance of cortical lesion detection for understanding and managing MS.