Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Oct 8:16:1678073.
doi: 10.3389/fneur.2025.1678073. eCollection 2025.

Evaluation of two AI techniques for the detection of new T2/FLAIR lesions in the follow-up of multiple sclerosis patients

Milica Mastilović  1   2 , Olivier Heinzlef  3   4 , Christian Federau  5   6   7 , Verónica Muñoz-Ramírez  8 , Marie Blanchere  3   4 , Jasmina Boban  2   9 , Francois Cotton  10   11 , Myriam Edjlali  1   12
Affiliations

Evaluation of two AI techniques for the detection of new T2/FLAIR lesions in the follow-up of multiple sclerosis patients

Milica Mastilović et al. Front Neurol. .

Abstract

Background: Multiple sclerosis is an inflammatory demyelinating disease of the CNS. Annual MRI exams are crucial for disease monitoring. Interpreting high T2/FLAIR lesion loads can be laborious. AI aids in lesion detection, and choosing between different solutions can be challenging.

Aim: This study compares two distinct software, Pixyl.Neuro.MS® and Jazz®, to assess their performance in T2/FLAIR lesion detection between two-time points.

Methods: Retrospective analysis included follow-up MRIs from 35 MS patients. Pixyl.Neuro.MS® automatically segments and classifies lesions. Jazz® automates the reading process and image display. Two readers (15 and 4 years of experience) conducted radiological analysis, followed by AI-assisted readings. A number of new lesions (NL) and reading times were recorded, with ground truth (GT) established by consensus. AI-detected lesions were classified as true (TP) and false positives (FP). Statistical analysis used SPSS (p < 0.05).

Results: Pixyl.Neuro.MS® readings averaged 2 min 46 s ± 1 min 4 s while using Jazz® 3 min 33 s ± 2 min 24 s. Over 50% of the population had a high lesion load (>20 lesions). Both software significantly improved NL detection (p < 0.01 for both), revealing them in more patients than standard readings. Standard reports found 8 NL in 2 patients, while AI-assisted readings detected at least 17 TP in 7 patients and rejected 61 FP lesions. GT detected 21 lesions in 19 patients.

Conclusion: Both AI software have been found to enhance NL detection in MS patients, outperforming standard methods. These tools offer crucial advantages for accurate disease monitoring.

Keywords: artificial intelligence; deep learning; lesion evaluation; magnetic resonance imaging; multiple sclerosis.

PubMed Disclaimer

Conflict of interest statement

Verónica Muños-Ramírez is an employee at Pixyl SA. Christian Federau is the founder and CEO of AI Medical AG. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The author(s) declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.

Figures

Figure 1
Figure 1
The figure represents the reading interface in Jazz® software. It allows a comparison of previous and new MRI FLAIR images of a patient with MS. The reader can easily switch from previous to new image, and vice versa, just by mouse clicking or using a keyboard shortcut (A), while there is as well lesion locking option (B) when the software automatically detects lesion’s anatomical location.
Figure 2
Figure 2
The figure shows representation of longitudinal evaluation of previous and new FLAIR images of a MS patient with high lesion load using Pixyl.Neuro.MS® segmentation mask. The lesions are color-coded in the segmentation mask: blue – stable lesion, red – new lesion. The white arrows in C3 indicate a new lesion detected using Pixyl.Neuro.MS®, confirmed by comparison of new (C2) and previous (C1) exams.
Figure 3
Figure 3
The figure shows an example of the lesion (indicated by the white arrow) missed by radiological evaluation by Jazz® software, while it was detected by AI-assisted radiological report made using Pixyl.Neuro.MS® software.
Figure 4
Figure 4
The figure represents a false positive lesion (indicated by the white arrow) which was segmented as a new lesion but actually represents an artifact located at the interface between cerebrospinal fluid and brain parenchyma.
See this image and copyright information in PMC

References

    1. Filippi M, Preziosa P, Banwell BL, Barkhof F, Ciccarelli O, de Stefano N, et al. Assessment of lesions on magnetic resonance imaging in multiple sclerosis: practical guidelines. Brain. (2019) 142:1858–75. doi: 10.1093/brain/awz144, PMID: - DOI - PMC - PubMed
    1. Thompson AJ, Banwell BL, Barkhof F, Carroll WM, Coetzee T, Comi G, et al. Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria. Lancet Neurol. (2018) 17:162–73. doi: 10.1016/S1474-4422(17)30470-2, PMID: - DOI - PubMed
    1. Baneke Peer. (2020). Atlas of MS 3rd Edition The Multiple Sclerosis International Federation. Available online at: https://www.msif.org/wp-content/uploads/2020/10/Atlas-3rd-Edition-Epidem... (Accessed June 25, 2025).
    1. Montalban X, Gold R, Thompson AJ, Otero-Romero S, Amato MP, Chandraratna D, et al. ECTRIMS/EAN guideline on the pharmacological treatment of people with multiple sclerosis. Mult Scler J. (2018) 24:96–120. doi: 10.1177/1352458517751049, PMID: - DOI - PubMed
    1. Baskaran AB, Grebenciucova E, Shoemaker T, Graham EL. Current updates on the diagnosis and management of multiple sclerosis for the general neurologist. J Clin Neurol. (2023) 19:217–29. doi: 10.3988/jcn.2022.0208, PMID: - DOI - PMC - PubMed

LinkOut - more resources