Non-invasive AI-powered Diagnostics by

How AI is used in Non-invasive Diagnostics.

cmashdown — 2024-04-01 22:41:06 UTC

AI is used for Non-invasive Diagnostics. Health data such as sleep patterns, heart rates and EKGs are collected through everyday devices such as smart watches, computers or cell phones. The data from these devices are then sent to cloud based applications where artificial intelligence algorithms analyze the data, identify patterns and determine if there is potential for underlying medical conditions.

In this particular article voice disorder detection is provided through the use of recording oneself remotely and the recordings are then analyzed by AI to detect underlying voice disorders such as vocal chord paralysis.

Benefits realized by the organization, clinicians, staff and/or patients through this specific use of AI.

cmashdown — 2024-04-01 22:41:37 UTC

Vocal disorders can greatly impact the quality of life for a patient. The voice detection allows for the AI to analyze the recordings done by patients and automatically detect patterns that are distinctive to particular disorders including cysts, polyps, spasmodic dysphonia and vocal cord paralysis.

By creating easier access to record their own data using everyday devices, it allows for inexpensive less invasive diagnosis for the patient and the facility. Allowing for a positive economic impact.

Through the implementation of this framework it allows for further growth in similar or related fields providing a greater technological impact for medical research and innovation.

The social impact is also great as it extends care to patients that may be located a further distance from the facility, cannot afford to receive care or may have disabilities that prevent them from accessing care.

https://ceur-ws.org/Vol-3651/HeDAI-3.pdf

cmashdown — 2024-04-01 22:41:51 UTC

Unfamiliar Term - Convolutional Neural Network (CNN)

cmashdown — 2024-04-01 22:42:04 UTC

A Convolutional Neural Network of CNN is an artificial neural network that has the ability to recognize patterns within images. CNN's can be used for classification purposes or feature identification. Some CNN's are even used for facial recognition.

Challenges

cmashdown — 2024-04-01 22:43:02 UTC

A significant challenge that would be data quality. Personal devices create a risk for data noise which can skew any data being analyzed. Skewed data can lead to misdiagnosis or incorrect user input.

Another problem that could be prevalent would be preservation of privacy. Patients may be hesitant to share personal information as it would be sent to a third party database and may be leary of data breaches.

cmashdown — 2024-04-01 22:44:30 UTC

cmashdown — 2024-04-01 22:46:13 UTC

cmashdown — 2024-04-01 22:47:31 UTC

cmashdown — 2024-04-08 22:16:18 UTC