Alvin Lang
April 11, 2025 22:36
The AI technology developed by the University of Florida uses a standard MRI scan to optimize Parkinson’s diagnosis to provide faster and more non -invasive detection and treatment.
According to NVIDIA’s report, a breakthrough AI drive tool is set to modify the diagnosis of Parkinson’s bottles using the standard MRI scan. This innovation can greatly improve the treatment and quality of life for patients by greatly rapidly detecting and treatment of Parkinson.
AI -based diagnostic tool
Developed by researchers from the University of Florida and the Leading Medical Center, this machine learning model can distinguish between Parkinson’s disease, multi -system atrophy (MSA) and gradual ultra -nuclear paralysis (PSP). These conditions often look similar in early brain scans, complicating diagnosis and causing treatment delays.
Michael S. OKUN, a senior author and medical advisor of Parkinson’s Foundation, emphasized the potential of AI to improve everyday diagnostic practices. Automated imaging differentiation techniques for Parkinsonism (AIDP) allow fast and accurate diagnosis without invasive procedures or professional scans.
Study and application
The result is published Zama neurologyAIDP shows that it can simplify early detection and treatment that matches the moon of the perception of Parkinson and the goals of the World Parkinson’s Day. The AI was trained using 645 brain scans, including images, previous studies and post -tests of new patients, and focuses on subtle brain tissue changes related to this disorder.
The main author, David Vaillancourt, pointed out the ability to detect a specific neurochromatic pattern, and worked as a pathological fingerprint for each condition. The training of this model used NVIDIA GPUs, including the NVIDIA Quadro P400, and with the TensorFlow library with NVIDIA CUDA, we achieved a 36 -hour training period.
Accuracy and the influence of the future
The AI showed 95%accuracy when diagnosing the case and surpassed the achievements of professional neurologists in challenging scenarios. In the post -mortem case, AIDP achieved 94%of the diagnosis confirmed compared to 82%accuracy by clinical diagnosis alone.
This high accuracy can greatly reduce misdiagnosis and alleviate the emotional burden of the patient and their family. Adaptability of AI tools across various hospitals and MRI machines suggests a wide range of possibilities from large hospitals to remote health services.
In addition, AIDP can improve clinical trials by ensuring the right patient selection, which is a continuous challenge in Parkinson’s research. This tool, which is licensed by NeuropACS, is expected to be integrated into the clinical environment when regulatory approval is obtained.
For more information on research, visit the NVIDIA blog.
Image Source: Shutter Stock