To develop reliable devices for DFU monitoring
The proposed devices combine passive infrared photodetectors with active illuminators. Specifically, we deliver (i) a passive HSI photo-detector, sensitive at NIR spectrum of 700nm 1000nm with an active tunable diode illuminator, operating at NIR spectrum, for optimizing reliability (in terms of sensitivity, specificity and accuracy) in detecting peripheral oxygen and tissue saturation- SpO2/StO2 and oxyhaemoglobin/deoxyhaemoglobin, at a spatial resolution of approximately 50pixels/cm, (ii) a passive Mid-IR photodetector, sensitive at spectrum 5.7μm-9.3μm with a Quantum Cascade Laser (QCL) optimized to capture additional tissue attributes such as elastin, collagen, lipid, amino amino-acids and carbohydrates necessary information for DFU early prediction and management at a resolution scale approximately of 10pixels/cm. and (iii) a thermal-IR sensing component capable of detecting hyperthermia/hypothermia distributions in ROIs with different levels of resolution for the PRO and In-Home version (full HD and 10 pixels/cm respectively).
Obj#2: Development of tuneable diodes (active illuminators), operating at 700nm-1000nm spectral range in order to increase reliability in detecting SpO2, StO2, HbO2 and Hb.
Current clinical studies have proven that NIR spectroscopy is capable of providing measurements of SpO2, StO2, HbO2 and Hb. However, the differences of the aforementioned values between a healthy and a non-healthy tissue are very small, especially at the early stages of the diseases. Therefore, very expensive HSI photodetector, operating at NIR, is needed. PHOOTONICS faces this problem, which causes cost-effectiveness issues, by combining active NIR illuminators, using tuneable diodes technology, with low-cost but optimized HSI photodetectors (Obj#1). In this way, we increase sensitivity, specificity and accuracy in detecting SpO2, StO2, HbO2 and Hb while keeping the overall cost small since dedicated infrared emission is accomplished.
PHOOTONICS (i) employs photonic enabled technologies targeted specifically for capturing key medical indicators for ulcer healing and monitoring, (ii) implements state of the art signal processing and machine learning algorithms to increase the discrimination accuracy (between healthy and non-healthy tissues) while maintaining hardware cost-benefit, (iii) develops a user-friendly interface in order to allow these devices to be operated by non-certified physicians, and even by patients (for the simplified In-Home version), and (iv) minimise operational cost in the monitoring and management of DFU by replacing invasive and costly practices with our non-invasive device and zero-consumables devices.
To develop cost-effective devices for DFU
Technical Innovation Objectives
Key Strategic Objectives
PHOOTONICS aims at developing innovative, reliable and cost-effective (in terms of almost zero operational cost and high return of investment) photonic-driven devices for Diabetic Foot Ulcers (DFU) monitoring and management which can be applied for wide use. The project supports two versions: (i) the PHOOTONICS In-Home, used for DFU monitoring by patients and (ii) the PHOOTONICS PRO operated by physicians at their premises.
Development of a silicon HSI photodetector optimized at specific NIR spectral bands (700nm-1000nm) capable of detecting SpO2, StO2, HbO2 and Hb, for early monitoring, prediction and management of DFUs.
Spectrometers are usually bulky and expensive. In this innovation, we exploit recent HSI research outcomes in silicon manufacturing to develop and optimize an HSI photodetector of high sensitivity in (700nm-1000nm) spectral bands and low sensitivity to the remaining spectrum wavelengths. The new photodetector will be of small size (of the size of 2-Euro coin) and will cost-effective compared to generic HSI sensors and competitive equipment. It will be also optimized to detect the spatial distribution of SpO2, StO2, HbO2, Hb in a 50pixels/cm accuracy. IMEC optimizes in-silicon HSI photodetectors in the NIR range.
Obj#3: Development of an optimized Thermal -IR sensor, capable of detecting spatial hyperthermia/hypothermia differences of two resolution levels for the In-Home and PRO version.
Conventional low-cost infrared (IR) sensing devices capture data of low discrimination accuracy and low spatial resolution which is not adequate for detecting small hyperthermia/hypothermia variations, which is considered as a crucial factor for early management of a diabetic foot. In this objective, we optimize thermal-IR sensors to yield high discrimination accuracy in detecting temperature variations in DFUs. We deliver two thermal-IR sensors; one for the PRO version of full HD spatial resolution and another for the In-Home version of 10pixels/cm resolution.
Obj#4: Development of Mid-IR photodetector, in the range of 5.7-9.3 μm combining with Quantum Cascade Laser (QCL) sources for increasing accuracy in detecting additional medical indicators, such as collagen, elastin, blood flow and water
The device will be based on a mid-IR photodetector combined with QCL structures that uses External Cavity setup based on broad gain QCL sources, to tune its transmission spectrum in a highly selective way, allowing for high specificity of DFU indicators detection and completely eliminating the need for optics (such as conventional Fourie Transform Infrared Spectroscopy - FTIR) between the light source and the light sensor. The plan is to cover the 5.7- 9.3 μm Mid-IR region with a maximum of 3 QCL sources and 9μm-13μm, based on the preliminary feasibility analysis of ALPES. The objective is to develop the optimum laser sources for this scope, while also integrate the sources in a miniaturised package, based on a MEMS cavity design of multiple lasers. In PHOOTONICS, QCL active illuminator technology is combined with Mid-IR photodetectors increasing their reliability. Particularly, the passive IR photodetectors detects temperature variations, maybe caused by external environmental factors. Active QCL illuminator captures additional medical indicators such as water, collagen, elastin and blood flow saturation, useful for the PHOOTONICS device to decide if the foot is on pathology or not.
Obj#5: Development of an embedded, adaptive signal processing and learning tool for to increase reliability, discrimination performance but also user friendliness
A major innovation of the PHOOTONICS device is the incorporation of an embedded, adaptive signal processing and learning framework. The objective of this embedded tool is twofold. First, it increases reliability and Associated with document Ref. Ares(2019)6059178 - 30/09/2019871908 - PHOOTONICS Part B Page 9 discrimination resolution of the new device by increasing SNR ratios (about 10db) and spatial resolution (more than 4x) incorporating advanced signal and image processing techniques in the area of (a) convolution processing, (b)
morphological analysis (non-linear convolution) and (c) Fourier/wavelet analysis. Second, it transforms the lowlevel captured data (by the sensors) to high-level medical indicators useful both for patients and physicians. Towards this direction, deep machine learning algorithms will be applied in the area of (a) image segmentation, (b) image enhancement (super-resolution), (c) classification (d) and regression. The developed signal processing and learning tool, apart from increasing reliability and discrimination levels of the new device (without increasing the hardware cost), will assist physicians in their diagnosis and therapeutic schemas or patients (for the in-home application case) in better management and daily handling the diabetic foot. The signal processing algorithms are embedded in PHOOTONICS hardware components, while the learning components are integrated into the respective software platform.
Obj#6: Development of the electronics and optics packaging interface to integrate the PHOOTONICS components into a single operational device.
The delivered photodetector and active illuminator will be integrated using appropriate electronics interfaces and optics control unit in order to deliver the single operational PHOOTONICS device. Two versions are delivered. The PHOOTONICS In-Home includes (a) the optimized IR sensing and (b) the embedded software component (adaptive signal processing and learning). We estimate the cost of this version at 900€. This version is suitable for patients. The PHOOTONICS PRO includes all photodetectors and active illuminators plus the software components, suitable for operating by physicians. We estimate the cost of this version at 60K€.
Obj#7: Execution of clinical studies to validate the reliability and cost-effectiveness of PHOOTONICS device
In the PHOOTONICS consortium, three big university hospitals are involved, to verify, validate and assess the reliability and the performance of the developed device for DFU monitoring, management and early diagnosis. The hospital clinics, participated in PHOOTONICS, have expertise in diabetic foot management. The clinical study includes diabetes type 1 and diabetes type 2. The patients are validated under a strict clinical framework supervised by the doctors. The goal is to assess the discrimination capabilities of the new device and derive factors that can prove that such a medical tool can improve management and diagnosis of the disease.