November 18, 2026
Health news, explained with evidence
LIVE

AI Tool Successfully Identifies Brain Microbleeds in Clinical Trial

🧠 Mind & Mental Health 2 min read 👁 4

A newly evaluated deep learning software has demonstrated high sensitivity and a low false-positive rate in detecting cerebral microbleeds on MRI scans, according to a recent clinical trial.

🧠Mind & Mental Health

Key points

  • Automated AI software evaluated for cerebral microbleed detection
  • Clinical trial tested model on 457 MRI examinations
  • Software achieved 88.35% sensitivity with 0.95 false positives per patient

Researchers evaluating the Neurophet SCALE MRI software have found that the deep learning-based tool successfully identifies cerebral microbleeds—small brain lesions that serve as important biomarkers for cognitive decline and cerebrovascular disease. Published in Nature Health Sciences, the retrospective single-center clinical trial tested the automated model on 457 MRI examinations involving 456 participants with a total of 515 microbleed lesions.

The software utilizes a two-stage deep learning pipeline, first generating lesion candidates using a multiscale backbone and convolutional Long Short-Term Memory, and subsequently reducing false positives through a classifier and brain tissue segmentation. According to the findings, the device exceeded its primary endpoints by achieving a lesion-wise sensitivity of 88.35 percent alongside a low false-positive rate of 0.95 per patient. Furthermore, the patient-wise sensitivity reached a perfect 1.0, with a specificity of 0.86.

While the study demonstrates promising diagnostic metrics, the retrospective and single-center design means that further validation by expert readers across broader populations is necessary to confirm its broader clinical utility. Several authors of the study are current or former employees of Neurophet Inc., the manufacturer of the software, and the trial was sponsored by the company.

What to know: Automated artificial intelligence tools may soon assist radiologists in identifying critical neurological markers like cerebral microbleeds, though patients should always consult qualified medical professionals regarding personal health decisions and neurological care.

Source Nature Health Sciences — nature.com/articles/s41598-026-77551-0 ↗

Contact the newsroom

Got a tip, a proposal or a partnership question? Write to us and we will reply shortly.