Dhonam Pemba

Our goal was to design, develop, and test a passive, portable, and inexpensive patient tracking system that measures the time between registration and treatment by an attending physician. We decided to use a Radio Frequency Identification (RFID) system to easily and transparently follow patients from the point of admittance to when they are seen by a doctor; this system incorporated a software program, in JAVA, as well as unobtrusive RFID “tags” carried on their person, to register patients and analyze their hospital activity. Data collected by our system would aid hospital management in making important decisions regarding staffing, resource allocation, and physical space available to the Emergency Departments. If this system was put into place in many hospitals nationwide, it would provide a large-scale basis of comparison of registration-to-treatment times across institutions, establishing a standard for evaluating managerial decisions. This would lead to smarter choices in hospital management, potentially serving to minimize wait times and reduce overcrowding.

The human brain undergoes significant changes in both its structural architecture and  function during the neuroanatomical course of human brain development. This article discusses reviews changes in brain anatomy (i.e., neuroanatomy) that occur during the developmental period in humans

This article discusses the developmental time windows for cognitive functions, i.e., critical periods during which the right stimuli can provide maximum return on developing cognitive function and its underlying neural circuitry. A comparison between child and adult neuroplasticity will be addressed.

The electrode serves as the critical component of Functional Electrical Stimulation, Brain Machine Interfaces and Neural Prosthesis.  The choice of electrode use in any stimulation system would depend on the application. In this Review, we examine various types of electrodes and analyze their sensitivity, invasiveness, reproducibility and effectiveness.

Neuroprostheses allow the possibility of restoring lost sensory and motor function by directly interfacing with the nervous system. This emerging field of neuroprosthetics aims to develop brain-machine and nerve-machine interfaces restore some of this lost neural function via selective electrical stimulation neural pathways. Here we provide background and on this emerging discipline.

With rise of deep learning thanks to gigantic scale of data generated by users on the Internet over the past 25 years as well as ever-growing computer processing power on the basis of Moore’s law reaching level of building neural network models surpassing other conventional statistical learning approaches, natural language processing has undergone rapid progress during the past 5 years, making chatbot capable of early childhood teaching and companionship possible. The biggest breakthrough when it comes to building chatbot for educational purposes would be Encoder-Decoder sequence to sequence conversational model with neural networks introduced first by Google Brain in 2014 [1], which is widely favored within academia projecting to revolutionize current applications of both goal-oriented chatbots and chitchat-bots. However, when trying to implement this pure data-driven approach into products, we ran into critical issues that cannot be solved with end to end approaches such as cohesive multi-turn dialogue, resulting in most successful production chatbots using hand crafted rules in accordance with Artificial Intelligence markup languages (AIML) instead. In this project, we propose a novel framework to utilize deep learning approach to generalize dialogues between lessons.