AI – Powered Software Enhances Parkinson’s Diagnosis with Over 96% Accuracy

Researchers at the University of Florida and the UF Health Norman Fixel Institute for Neurological Diseases have created a new AI software designed to assist clinicians in differentially diagnosing Parkinson’s disease and related conditions. This innovation aims to reduce diagnostic time and enhance accuracy, achieving precision levels exceeding 96%.

Parkinson’s Diagnosis Faces Accuracy Challenges and High Misdiagnosis Rates

Research shows that the accuracy of Parkinson’s disease (PD) diagnoses ranges from 55% to 78% during the first five years of assessment. This is partly due to the similarities between Parkinson’s disease and related movement disorders, which can complicate the initial diagnosis.

PD includes a range of conditions, from the most common idiopathic Parkinson’s to other movement disorders like multiple system atrophy Parkinsonian variant and progressive supranuclear palsy. These conditions share motor and nonmotor symptoms, such as gait changes, but have distinct pathologies and prognoses. Studies suggest that approximately one in four to one in two patients may receive a misdiagnosis.

“In many cases, MRI manufacturers don’t communicate with each other due to marketplace competition,” said David Vaillancourt, Ph.D., chair and a professor in the UF Department of Applied Physiology and Kinesiology. “They all have their own software and their own sequences. Here, we’ve developed novel software that works across all of them.”

Additionally, Vaillancourt stated that while there is no replacement for the human aspect of diagnosis, even the most experienced physicians specializing in movement disorders can benefit from a tool that enhances diagnostic accuracy across different conditions.

Model Shows High Accuracy in Differentiating PD, MSA, and PSP, Consistent Across Tests and Patient Variables

The Automated Imaging Differentiation for Parkinsonism (AIDP) software is an MRI processing and machine learning tool that employs a noninvasive biomarker technique. It utilizes diffusion-weighted MRI to assess water molecule movement in the brain, allowing the team to detect areas of neurodegeneration. The machine learning algorithm, validated against clinical diagnoses, analyzes the MRI results and provides clinicians with information to distinguish between different types of Parkinson’s disease.

The study included 249 patients from 21 sites, with 99 diagnosed with Parkinson’s disease (PD), 53 with multiple system atrophy (MSA), and 97 with progressive supranuclear palsy (PSP). The AI software trained on an additional dataset of 396 patients, with 78% of cases used for development and 22% reserved for testing. Over three years, researchers collected brain samples from 49 patients (5 with PD, 5 with MSA, and 39 with PSP) for further analysis.

The model demonstrated strong accuracy in distinguishing between PD and related conditions like MSA and PSP, with high scores for sensitivity and specificity. The model consistently performed well, even when retested with new scans or excluding age and sex. Additionally, the model reliably identified patients across various testing sets. More severe symptoms in MSA patients linked to more accurate diagnoses, while symptom severity had little effect on PSP diagnoses.

“This is an instance where the innovation between technology and artificial intelligence has been proven to enhance diagnostic precision, allowing us the opportunity to further improve treatment for patients with Parkinson’s disease,” said Michael Okun, M.D., medical adviser to the Parkinson’s Foundation and director of the Norman Fixel Institute for Neurological Diseases at UF Health.

Parkinson’s Disease Prevalence Has Doubled in 25 Years, With Over 8.5 Million Affected

According to WHO statistics, the prevalence of PD has doubled over the past 25 years. Global estimates from 2019 reported over 8.5 million individuals living with PD. Additionally, PD caused 5.8 million disability-adjusted life years (DALYs) in 2019, reflecting an 81% increase since 2000, and led to 329,000 deaths, more than double the number reported in 2000.

There are currently no specific tests available to diagnose Parkinson’s disease. A neurologist makes the diagnosis based on the patient’s medical history, symptom review, and a neurological and physical examination. Diagnosing Parkinson’s disease can be a gradual process. Healthcare professionals may suggest regular follow-up appointments with neurologists who specialize in movement disorders to monitor symptoms and assess the condition over time to confirm the diagnosis.

The introduction of this new AI model aims to streamline the diagnostic process for Parkinson’s disease and related conditions. By providing a highly accurate, non-invasive tool for distinguishing between Parkinson’s disease, MSA, and PSP, the model has the potential to reduce diagnostic uncertainty and speed up decision-making. According to researchers, it could provide more timely and precise diagnoses, allowing for earlier intervention and better management of these conditions. 

Related posts:

Autistic Individuals at Triple Risk for Parkinson’s-Like Symptoms, Study Finds First-of-its-kind AI Model Uses Brain Scans to Reveal Link Between Aging and Cognitive Decline Neurosurgery Now: NVIDIA’s Holoscan AI Brings Real-Time 3D Vision to Enhance Surgical Precision at 2025 GTC Palantir Stock is Skyrocketing as its AI Tech Promises Big Gains

Author

Denisse Sandoval