A PHP Error was encountered

Severity: 8192

Message: Return type of LayerShifter\TLDExtract\Result::offsetExists($offset) should either be compatible with ArrayAccess::offsetExists(mixed $offset): bool, or the #[\ReturnTypeWillChange] attribute should be used to temporarily suppress the notice

Filename: src/Result.php

Line Number: 189

Backtrace:

File: /home/u149759368/domains/techthop.com/public_html/application/third_party/domain-parser/layershifter/tld-extract/src/Result.php
Line: 27
Function: _error_handler

File: /home/u149759368/domains/techthop.com/public_html/application/third_party/domain-parser/composer/ClassLoader.php
Line: 444
Function: include

File: /home/u149759368/domains/techthop.com/public_html/application/third_party/domain-parser/composer/ClassLoader.php
Line: 322
Function: Composer\Autoload\includeFile

File: /home/u149759368/domains/techthop.com/public_html/application/third_party/domain-parser/layershifter/tld-extract/src/Extract.php
Line: 167
Function: loadClass

File: /home/u149759368/domains/techthop.com/public_html/application/third_party/domain-parser/layershifter/tld-extract/src/static.php
Line: 35
Function: parse

File: /home/u149759368/domains/techthop.com/public_html/index.php
Line: 331
Function: require_once

The Best Dementia Care: AI Diagnoses Dementia Correctly - TechThop

The Best Dementia Care: AI Diagnoses Dementia Correctly

  • Facebook
  • Twitter

The Best Dementia Care: AI Diagnoses Dementia Correctly

As a result of improvements in public health over the last few decades, more people are living into old age around the world. As a result, dementia, including Alzheimer's disease, and other conditions associated with aging are on the rise.

In light of a predicted physician shortage in the next decades, this might impede prompt treatment for individuals in need. The Boston University School of Medicine recently reported that computational techniques may help ease some of the challenges associated with delivering dementia care to the aging population.

“The use of automation to support diagnosis would help doctors and patients plan treatment accordingly, even when a neurology or neuroradiology specialized physician is overloaded. The corresponding author Vijaya B. Kolachalama, Ph.D., FAHA, is an assistant professor at BUSM. 

The data from previous research shows that AI models can choose between "disease" and "no disease" in a straightforward manner, but that is not how clinicians treat patients. Instead, they should consider all possible conditions affecting a patient in their clinic, depending on physical examination, neuropsychological testing, laboratory results, and imaging to establish a distinctive signature that solidifies the diagnosis.

Since it allows a computer to zero in on the exact cause of a patient's disease even when other possible causes exist, Kolachalama considers this research to be more in line with "real world" situations. The model can achieve this when a broad differential diagnosis of possible illnesses is provided.

Alzheimer's disease is the most common cause of dementia as we know it, but chronic alterations in a person's mental state can also occur with Parkinson's disease, geriatric depression, nutritional deficiencies, and more. "With our study, we demonstrate a computational approach for providing accurate diagnoses during this diverse landscape of neurologic diseases," he explains.

A variety of computer models were developed to process data collected during a typical work-up of a dementia patient, including neuro-psychological and functional testing, medical history, physical examination, demographics, and MRI scans. From this vast set of inputs, a neural network was trained to extract disease-specific signatures.

By using specialized methods in machine learning, they identified the precise pieces of data that their model used in its diagnostic decision-making, including important neuropsychological test scores, laboratory values, and physical examination findings that could be associated with a specific disease.

As a result, they applied these methods to localize dementia-related changes in MRI scans and found that the locations marked as "important" by the model corresponded to brain regions with microscopic signs of degeneration. The AI models were also compared with a group of international physicians. Experts and models were presented with the same set of patients and asked to provide diagnoses based on the same information.

There was no difference in accuracy between the doctors and the computer. The use of computational strategies can alleviate some of the difficulties associated with dementia care for an aging population, Kolachalama argues. When patients are unable to obtain specialized neurologic care, our work can help fill in the gaps and provide timely information about their health and loved ones' wellbeing.

More Tech