Aging Brain - Dementia

Value Proposition

The number of people surviving into their 80s and beyond is expected to grow dramatically and it is expected to contain up to 20% of the total population by 2030. As the number of older people grows rapidly, neurodegenerative diseases and brain diseases associated with aging are also expected to rise significantly.  It is obvious that there will be a significant increase in imaging of the aging brain. The goal of this use case is to collect a large data set of brain imaging from healthy old population and patients with dementia. This use case then can be used to train different AI algorithms to detect different patterns of brain atrophy which can be helpful for diagnosis the underlying etiology of dementia. 

Most grading systems of brain volume loss are subjective with low inter and intra observer agreement. The value of this use case is to train AI models to detect the brain atrophy, calculate the brain volume, calculate the volume of each brain structure , determine the atrophy grading system and suggest the possible cause of dementia.

Narrative(s)

 An 80 year old man presents to the neurology clinic with slow onset cognitive decline, imbalance and abnormal gait. As a work up the neurologist orders the brain MRI with dementia protocol. The general radiologist who reports this MRI detects mild general volume loss and grade I atrophy of hippocampi but misses atrophy of midbrain. The diagnosis of progressive supra-nuclear palsy is missed in this patient because it is a rare condition and the radiologist has limited experience in this field. 

Workflow Description

To collect the high resolution T1 and FLAIR brain images in healthy old population as well as patients with different types of dementia. The images will be annotated by expert neuroradiologists regarding different grades of regional brain atrophy and presence of typical imaging pictures of dementia. In addition, the total brain volume,  volume of CSF and ventricles and volume of different brain structures such as hippocampi are documented.

Age: varied

Sex at birth: male, femal

Global cortical atrophy (GCA) scale: varied

Medial Temporal atrophy score: varied

Posterior Parietal Atrophy score ( Koedman Score): varied

Presence of Frontotemporal atrophy: absent, present

Presence of midbrain atrophy: absent, present

Presence of Caudate head atrophy: absent, present

Presence of cerebellar atrophy: absent, present

Estimation of Total brain Volume and comparison with normal cohort: varied

Estimation of CSF volume and comparison with normal cohort: varied

Estimation of Ventricular system volume and comparison with normal cohort: varied

Estimation of Volume of especial brain structures such as hippocampi, frontal lobes, temporal lobes, etc and comparison with normal cohort: varied

Fzekas Scale (from FLAIR sequence): varied

Inputs

DICOM Study

Primary Outputs

Global Cortical atrophy (GCA) scale

Posterior Parietal Atrophy score (Koedman Score)

Hippocampi volume

Temporal lobes volume

• Besides improving the accuracy of radiology reports, this use case can help the researcher to design better clinical trials by selecting similar patients.

No known related public datasets at this time,  please alert us if you know of any.