3D Camera Data Fusion for AI-Based Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing non-invasive glucose monitoring systems face challenges in sensitivity, selectivity, and repeatability, while current dental caries detection methods are subjective and x-ray imaging poses health risks, and there is a need for early detection of breast cancer without ionizing radiation.
Innovation Solution
Utilizing fiber-based super-continuum lasers and near-infrared or short-wave infrared light sources for non-invasive glucose monitoring, dental caries detection, and breast cancer imaging, combined with multi-modal generative artificial intelligence models for enhanced accuracy and early detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If invasive blood draws are used for glucose monitoring, then measurement precision is improved, but ease of operation deteriorates due to pain and inconvenience
Solution Approach 1:
The patent replaces the mechanical invasive blood draw system with an optical measurement system. Near-infrared light is used to penetrate tissue and detect glucose levels non-invasively, substituting the mechanical needle-based system with a painless optical sensing approach that maintains measurement capability while eliminating the harmful mechanical intrusion.
Solution Approach 2:
The patent introduces near-infrared light as an intermediary medium to access blood constituents without direct contact. The light acts as a mediator that can penetrate through tissue (skin, teeth) to reach blood components, enabling indirect measurement of glucose levels without requiring physical blood extraction.
2Ease of operation
If non-invasive optical methods are used, then ease of operation is improved, but measurement precision deteriorates due to insufficient signal levels from blood constituents
Solution Approach 1:
The patent changes the wavelength parameter of the light source to the near-infrared region (700-2500 nm), which has optimal penetration depth through biological tissues and corresponds to absorption characteristics of blood constituents. This parameter change enables both non-invasive operation and sufficient signal detection by matching the optical properties of tissue and blood components.
Solution Approach 2:
The patent employs pulsed light sources that emit near-infrared light in periodic pulses rather than continuous illumination. This periodic action allows for time-resolved detection, improving signal-to-noise ratio by synchronizing detection with the light pulses and enabling discrimination between scattered and non-scattered photons, thereby enhancing measurement precision while maintaining non-invasive operation.
3Ease of operation
If skin is used as the measurement site, then ease of operation is improved, but measurement precision deteriorates due to spectral artifacts from skin
Solution Approach 1:
The patent extracts or removes the skin from the measurement pathway by utilizing teeth as an alternative access point. Teeth provide a different optical pathway with fewer interfering spectral artifacts from skin components, allowing the near-infrared light to reach blood constituents with minimal interference, thereby improving spectral selectivity while maintaining non-invasive accessibility.
4Measurement precision
If brighter light sources are used to increase signal levels, then measurement precision is improved, but use of energy deteriorates
Solution Approach 1:
The patent uses pulsed light sources instead of continuous illumination, emitting near-infrared light in periodic pulses. This periodic action reduces average power consumption compared to continuous operation while maintaining sufficient peak signal levels for precise detection. The pulsed regime allows the detector to integrate signals during the pulse and reject noise between pulses, improving signal-to-noise ratio without proportionally increasing energy use.
Solution Approach 2:
The patent employs rapid pulsing at high repetition rates (10 MHz to 1 GHz) to maintain continuous measurement capability. Although individual pulses are brief, the high repetition rate ensures continuous sampling and data acquisition, maintaining the continuity of useful measurement action while using low average power due to the duty cycle of pulsed operation.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Achieves accurate, non-invasive glucose monitoring, early dental caries detection, and effective breast cancer screening without ionizing radiation, leveraging advanced light sources and AI for improved sensitivity and specificity.
Implementation Method 1
an array of laser diodes configured to generate a light having an initial light intensity and one or more optical wavelengths
Implementation Method 2
The detection system is configured to perform a time-of-flight measurement based on a time difference between a first time in which the array of laser diodes generate light and a second time in which the photodiode array receives a received portion of light reflected from the object
Implementation Method 3
a detection system comprising a photodiode array and further comprises at least one second lens and one or more spectral filters in front of at least a part of the photodiode array
Implementation Method 4
The use of near-infrared spectroscopy with brighter light sources such as fiber-based supercontinuum lasers, super-luminescent laser diodes, light-emitting diodes, or multiple laser diodes to increase near-infrared signal levels from blood constituents
Implementation Method 5
pattern matching in spectral fingerprinting and software techniques to identify blood constituents through the teeth
Data Source
AI summary
3D cameras may serve as an input to a multi-modal generative artificial intelligence (GAI) model operating on a processor coupled to a non-transitory computer readable medium. Examples of 3D cameras include direct or indirect time-of-flight sensors or structured light systems and may also be coupled to 2D cameras. The GAI comprises a vision transformer configured to analyze an item in an input video or image. The vision transformer comprises self-attention and positional encoding layers. The GAI may be trained using reinforcement learning or fine-tuning involving training images or videos, and it may perform data fusion by combining the 3D information with data from other sensors. The GAI may perform anomalous occurrence detection by training on images or videos corresponding to normal occurrences. The GAI detects differences in an image or video that fall outside of a threshold value. The GAI may also provide safeguards for privacy issues.