The research, led by Dr. Zeynep Bendella and colleagues, involved a retrospective analysis of 123 subjects, including patients diagnosed with iNPH, AD, as well as age- and sex-matched healthy controls. The team employed advanced AI software, mdbrain v4.4.1, to automatically determine various brain areas and ventricular volumes. The tool provides precise, volumetric measurements that were then compared with traditional semi-quantitative markers such as the Evans' index and corpus callosal angle.

Key Findings:

• Significant Volumetric Differences: The study revealed substantial differences in brain areas and ventricular volumes among the groups. Notably, iNPH patients exhibited a pronounced increase in total ventricular volume (+67%) and specific enlargement in the lateral (+68%), third (+38%), and fourth ventricles (+31%) compared to controls. This contrasted with the global gray matter reductions observed in AD patients.
• Distinctive Structural Changes: AI-based volumetry detected global ventriculomegaly and marked white matter reduction in iNPH patients, particularly in the supratentorial regions, with relative preservation of gray matter. This structural signature was distinct from the atrophic patterns typically seen in AD.
• Enhanced Diagnostic Accuracy: The integration of AI volumetry with established radiologic markers demonstrated the potential to enhance diagnostic accuracy for iNPH, differentiating it from conditions with overlapping symptoms, such as AD.

Andreas Lemke, CEO of mediaire, highlighted the significance of these findings: "The ability to accurately differentiate iNPH from other neurodegenerative diseases is critical for patient management. The study demonstrates that AI-based MRI volumetry can offer a more detailed and reliable assessment of brain structures, paving the way for better diagnostic clarity and tailored treatment strategies."

The application of AI in this context not only provides a more objective and quantitative analysis of brain alterations but also potentially improves the identification of candidates who may benefit from therapeutic interventions, such as ventricular shunting. This advancement is particularly important given the overlap in clinical presentations between iNPH and AD, which has historically led to diagnostic challenges.

The research team calls for further studies to validate these findings across larger and more diverse populations, with the aim of integrating AI-based tools into routine clinical practice. As the technology continues to evolve, it holds promise for transforming the landscape of neurodegenerative disease diagnostics and management.