Non-Invasive Ultrasound-Based Tool Accurately Detects Infant Meningitis

Even when treated, it may leave lasting damage, such as cognitive impairment or neurological issues. Despite advancements in healthcare, meningitis continues to pose a significant threat to child health, particularly in low- and middle-income countries where access to early diagnostic tools is limited. Currently, diagnosis relies on lumbar puncture to collect and analyze cerebrospinal fluid—an invasive procedure that carries risks and limitations. In high-income settings, this method often yields low diagnostic value due to routine overuse, while in low-resource areas, its underuse leads to widespread underdiagnosis and empirical treatment errors. Now, a new ultrasound-based device that can detect infant meningitis with high accuracy offers a promising, non-invasive alternative to conventional diagnostic methods.

In an international study, led by the Barcelona Institute for Global Health (ISGlobal, Barcelona, Spain), researchers tested the unique NEOSONICS ultrasound-based cell counting device that enables fast, easy, and non-invasive screening of infant meningitis. The device uses high-frequency ultrasound applied through the open fontanelle of an infant's skull to visualize and analyze cerebrospinal fluid. A deep learning algorithm processes the images to detect and count cells, identifying signs of inflammation indicative of meningitis. Its cost-effective, portable, and easy-to-use design makes it suitable for settings where lumbar puncture is contraindicated or impractical. By reducing the need for invasive testing, the device has the potential to minimize unnecessary antibiotic use, prevent complications, and support non-invasive monitoring of treatment.

The device was validated through a clinical study conducted between 2020 and 2023, involving more than 200 infants aged up to 24 months. The findings, published in Pediatric Research, show that the device accurately classified 17 of 18 confirmed meningitis cases and 55 of 58 non-meningitis controls, achieving approximately 94% sensitivity and 95% specificity in detecting high white blood cell levels in cerebrospinal fluid. Moving forward, researchers aim to further integrate artificial intelligence to improve diagnostic precision and promote broader adoption of the device in clinical practice.

"Introducing a non-invasive tool could reduce unnecessary antibiotic use, prevent complications associated with lumbar puncture, and improve both early diagnosis and non-invasive monitoring of treatment response," said Quique Bassat, Director General of ISGlobal, ICREA researcher, and senior author of the study.

Related Links:
ISGlobal