Amr Elsawy, Mohamed Abou Shousha, Mohamed Abdel-Mottaleb Electrical and Computer Engineering Dept, Ophthalmolgy Dept., Bascom Palmer Eye Institute, Electrical and Computer Engineering Dept Optical Coherence Tomography (OCT) is a technology that is convenient for imaging the eye because it is non-contact and no-invasive. High resolution images of the anterior and posterior segments of the eye can be obtained using OCT. The anterior segment involves the cornea and its microlayers. Measuring the thickness of different corneal microlayers, from OCT images, is important for diagnosing several corneal diseases such as Fuchs dystrophy, keratoconus and corneal graft rejection. Segmentation of corneal microlayers is needed to be able to measure the thickness. However, manual segmentation of those microlayers in OCT images is subjective and time consuming. Therefore, automatic segmentation is a necessity. Many methods were proposed for segmentation of corneal microlayers on OCT images, but none of these methods segment all the microlayers and they are not robust. Moreover, there is no large annotated database of corneal OCT images which prevents the application of machine learning methods such as deep learning in the segmentation. We presented new corneal OCT image segmentation methods using polynomial fitting, Randomized Hough Transform (RHT) or graph search. Our proposed methods can segment up to six corneal microlayers. The proposed methods help to measure thickness of corneal microlayers to quantify the diseases associated with those layers. The proposed methods were validated against manual segmentation of a random sample of images. The manual segmentation is done by manual trained operators. The segmentation error between the automatic segmentation and the manual segmentation is comparable to the inter-operator error between the manual operators and the significance test shows that there is no significance between them. Then, the segmentation is carried out along the cuts of a radial scan to provide 3D surfaces of corneal microlayers. Then, the surfaces are corrected due to light refraction and the thickness maps of corneal microlayers are obtained by measuring the thickness between those surfaces.
Elikem K. Tettey-Tamaklo, Solomon Mensah Industrial Engineering Respiratory distress syndrome (RDS) is a lung disorder that affects the normal breathing of preterm infants by preventing their lungs from staying open leading to obstructed breathing. This disorder is responsible for much of the largely preventable preterm infant mortality in many developing countries. The solution to this disease would be to provide the preterm babies with a constant supply of oxygen until their lungs mature enough to stay open without help. Commercial respiratory assistants are very expensive and made for first world facilities, but are being used in third world hospitals. Their combination of high cost and complexity result in the inability of local hospitals to afford or maintain them when they break. The result is less of these devices available to cater to the growing infant mortality crisis within developing countries. In an attempt to curb the problem, a cost-effective and practical intervention would be necessary. This intervention comes in the form of a completely redesigned bCPAP. This device is a respiratory device that provides the ideal mixture of oxygen and air, delivered to the delicate lungs of the baby. Therapeutic Innovations redesigns medical devices towards third world countries by removing unnecessary complex features while still maintaining reliable functionality. Part of our solution is to use pre-fabricated parts from medical suppliers and assembling them in a novel modular fashion to create our bCPAP. This removes manufacturing and approval costs, and allow us to focus on designing our product towards the conditions of third world countries.
Aaron A. Stock, Vita Manzoli, Diana Velluto, Felicia Pagliuca, Alice A. Tomei Biomedical Engineering Type 1 diabetes is an autoimmune disease characterized by T cell mediated destruction of the β cells of the islets of Langerhans in the pancreas. Pursuant to the loss of these insulin secreting cells, blood glucose cannot be regulated and those affected suffer reduced quality of life or even mortality. Islet transplantation presents a promising opportunity for curing this disease. However, this procedure is only indicated for adults with the most severe and unnoticed hypoglycemia due to the required systemic immunosuppression and limited availability of donor islets. Stem cell-derived insulin secreting cells (hereafter, scβ cell clusters) represent a potentially unlimited source of islets for transplant, and in combination with conformal encapsulation, may be the first immune-suppression free islet transplant platform that is indicated for the majority of people affected by autoimmune diabetes.
W.M. Batchelor1, Noel Ziebarth1, Vincent Moy2 1Department of Biomedical Engineering, College of Engineering, University of Miami 2Department of Cellular Physiology and Molecular Biophysics, Miller School of Medicine, University of Miami Glaucoma is a group of several related eye diseases that lead to progressive optic nerve degeneration and retinal ganglion cell (RGC) death. As of today, over 60 million people around the world suffer from some degree of permanent vision loss due to glaucoma, making glaucoma globally the most common cause of permanent visual impairment. Because of this, researchers and clinicians have focused on developing novel treatment strategies for glaucoma; however, many forms of glaucoma are difficult to treat and will eventually result in blindness. Glaucoma is a very complex disease, and animal models do not accurately capture all facets of its extremely complex pathogenesis. Lack of an appropriate glaucoma model slows development of a suitable treatment. An in vitro model of glaucoma, which better recapitulates the development of glaucoma in humans than traditional animal models, would be very useful in both studying the disease and creating treatments for it. This project proposes a novel in vitro model of human glaucoma that will incorporate both stem-cell derived retinal ganglion cells and unique biomaterials. For the first time, atomic force microscope (AFM) based methods of nanolithography will be used to chemically, physically, and topologically functionalize an electrospun collagen scaffold to form the base of the model. Over the course of this project, we will design and build a customized AFM for nanolithography on biomaterials, use this AFM to functionalize a biomaterial scaffold, combine cells to the scaffold to assemble the model, and finally, we will validate our model to determine how effective it is at simulating glaucoma.
Siobhan Williams1,2, Giovanni Gregori3, Marco Ruggeri1,2, Yu-Cherng Chang1,2, Florence Cabot1,3, Arthur Ho1,2,4, Sonia Yoo1-3, Jean-Marie Parel1-4, Fabrice Manns1,2 1Ophthalmic Biophysics Center, University of Miami Miller School of Medicine, FL; 2Biomedical Optics and Laser Laboratory, Department of Biomedical Engineering, University of Miami, Coral Gables, FL; 3Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL; 4Vision Cooperative Research Centre, Sydney, Brien Holden Vision Institute, UNSW, Sydney, Australia Dynamic mechanisms of the eye’s accommodative apparatus are a key component in understanding presbyopia. A recently developed extended-depth spectral domain optical coherence tomography (SD-OCT) imaging system provides a non-invasive technique to stimulate accommodation and capture the eye’s response (Ruggeri et al. 2012). Large datasets generated by this approach require an automated and accurate method to quantify the eye shape using OCT images. This project introduces a fast, automated algorithm for quantifying dynamic changes of the crystalline lens during accommodation using two-dimensional OCT images. The algorithm attains a delicate balance between the inevitable trade-off between robustness and speed to efficiently process large volumes of data.
To develop a system that can simultaneously measure the power of the eye and changes in the lens shape combining an adjustable accommodation stimulus, an autorefractor, and an Optical Coherence Tomography (OCT) system