Multispectral imaging and associated computational algorithm for functional characterization of ocular pathologies

Retinal pathologies are major cause of visual impairment and can be partly prevented through periodic check-up and monitoring. Different risk factors like: diabetes, formation of atherosclerosis, bleeding of vessels, retinal angiomatous proliferation, blockage of the retinal circulation due to fat or blood clotting etc. hinder blood flow in retinal circulation by narrowing vessel structures. In retinal neovascularisation, abnormal growth of tiny, leaky blood vessels causes alternative medium of blood flow. Often it may lead to vitreous haemorrhage and neovascular glaucoma, impairing natural pressure relieving mechanisms of eye. Other hand, ROP is becoming a serious concern despite of significant research in neonatology. It is a serious vasoproliferative disorder that affects the retina of premature infants and can manifest to severe conditions like neovascularisation and progress to retinal detachment causing permanent visual impairment.This proposal aims at developing technology for patient comfort centric ocular imaging by translating fundamental scientific discoveries in bio-photonic imaging associated with characteristic response of anterior and posterior chambers of eye at different wavelengths of light illumination into specific clinical interventions which can further provide disease specific quantitative biomarker at cellular and vascular level. Our focus is to develop affordable (low cost), scalable, point-of-care and frugal technological solution to empower existing clinical set-up for ocular imaging in health and pathology. In this project, we aim to develop a hardware-software co-designed portable imaging module with associated computational mechanisms to assist clinicians perform fast and accurate diagnosis of retinal diseases e.g. diabetic retinopathy (DR), glaucoma, retinal neovascularization, retinopathy of prematurity (ROP) etc. Coupled with the special computational hardware device, the imaging module can be used by semi-skilled paramedics working at rural and primary healthcare centers for fast and high-precision screening ofocularconditions in health and pathology.

An ultra-low-powerMSP430-based embedded system and associated computational algorithm to develop a point-of-care cardiac pre-screening device

The system is a point-of-care (POC) device that can deliver heart-care services to the rural population and bridge the rural–urban divide in healthcare delivery.Normal heart sounds provide an indication of the general state of the heart in terms of rhythm and contractility, while changing its signature due to any cardiovascular pathology for additional murmurs that contributesignificant information in diagnosis. Eventually, the presence of lung sound affects the signature characteristics of heart sound when mixed with the later one in spatio-temporal domain, further separation of which becomes difficult through normal filtering approaches. This work aims todetermine a framework for utilization of automated HS analysissystem for community healthcare and healthcare inclusion. It focuses on the development of a standalone system using a TI-MSP430, aiming to acquire and filter the heart sound signal based on modern signal processing algorithms.Initially, the acquired sound waves are passed through a band pass filter bank followed by an amplification unit, enabling elimination of high frequency noise signals and subsequent amplification of the signal of interest. When captured at MSP430 terminal, the microcontroller converts the analog signal into its digital counterpart, maintaining Nyquist rate and stores the time-sequence data into its on-board memory blocks. The computational algorithm, based on adaptive and wavelet signal processing techniques, integrated on the hardware platform further process the acquired data to reduce the effect of lung sound from the composite signal through various statistical measures. Output of the MSP430 system can be either fed to a digital display or microphone system for accurate determination of the heart sounds which can provide diagnostic clinical inference about the heart in health and pathology.

Non-invasive andin vivo dynamic imaging of microvasculature and tissue perfusion using laserspeckle contrast technique

A relative measurement of blood flow changes is of major importance in the clinical domain for quantifying vascular change due to tissue injury or diseased condition. In superficial organs (viz. eye, skin), changes in blood flow are caused by different risk factors: atherosclerosis, blockage due to abnormal deposition of extracellular matrix within blood vessels, high cholesterol profile, diabetes and other inflammatory disorders like burn injury, wounds, as well as growth of benign and malignant tumours (lesions). These pathological conditions cause improper supply of oxygen and nutrients to the cells, affecting the functional and structural integrity of vasculature. In this context, Laser speckle contrast imaging (LSCI) provides a low-cost solution with high resolution probing facility among blood flow imaging modalities like: laser Doppler flowmetry, Doppler optical coherence tomography, polarization spectroscopy, photo-acoustic tomography, magnetic resonance imaging etc. This work presents speckle contrast imaging technique for in vivo imaging of microvasculature and blood perfusion in different physiological conditions. The objective is to develop computational algorithms for processing speckle images and evaluates its efficacy for quantifying changes in blood flow during abnormal state. This probing mechanism can be utilized for different applications like: detection and tracking of emboli in vascular model, changes in blood perfusion in skin flaps, changes in retinal blood perfusion due to external stimuli or various ocular pathologies, non-invasive and label free retinal angiography for the pathologies aggravating neovascularization, functional characterization of various cutaneous woundsand their stages of progression for periodic healing study etc.