STOCHASTIC DECISION MAKING FOR ADAPTIVE CROWDSOURCING IN MEDICAL BIG-DATA PLATFORMS

STOCHASTIC DECISION MAKING FOR ADAPTIVE CROWDSOURCING IN MEDICAL BIG-DATA PLATFORMS

ABSTRACT

Two novel algorithms for adaptive crowdsourcing in medical imaging big-data platforms is considered, namely, a max-weight scheduling algorithm for medical cloud platforms and a stochastic decision-making algorithm for distributed power-and-latency-aware dynamic buffer management in medical devices. In the first algorithm, medical cloud platforms perform a joint queue-backlog and rate-aware scheduling decisions for matching deployed access points (APs) and medical users where APs are eventually connected to medical clouds. In the second algorithm, each scheduled medical device computes the amounts of power allocation to upload its own medical data to medical big-data clouds with stochastic decision making considering joint energy-efficiency and buffer stability optimization.

INTRODUCTION

In recent years, the volume of medical data generated by large hospitals is becoming increasingly large due to technological advancements in medical devices, including high-resolution magnetic resonance imaging (MRI), motion MRI, ultrasound, and digital microscopy. Furthermore, centralized storage of medical records is a common practice for sharing medical data among medical practitioners. Oftentimes, medical records are collected and uploaded to the centralized medical record using modern mobile equipments, such as smart phones, and via wireless access points (APs). Because of the sensitive nature of medical data, data aggregation, needs to be privacy preserving. Therefore, interconnecting medical storage platforms with external networks (such as the Internet) is not recommended. Medical data in the proposed medical storage platform is often gathered and organized by fixed users—e.g., purposed medical tablets, smartphones, computed tomography scanners, etc.—with the principle of crowdsourcing.

EXISTING SYSTEM

Ø There is no prior work on scheduling and buffer management in the context of medical big-data platforms

Ø In the general wireless scheduling literature, the sum-rate-maximization (SRM) scheduling is one of the well known schemes that is most closely related to proposed max-weight scheduling.

Disadvantages

Ø Since the SRM schedules user based only on data rates

Ø SRM Schedules has no effect on buffer-backlog and management.

PROPOSED SYSTEM

Ø In this proposed medical storage system, 60-GHz wireless technologies is considered for in-hospital wireless network access. The choice of wireless technologies has been widely advocated and accepted in the literature because of high data rates achieved by ultrawide-bandwidth; e.g., 2.16 GHz in one subchannel and four subchannels in one channel.

Ø Two algorithms are proposed to address the scheduler and buffer management in medical platforms.

Ø The proposed medical platform makes scheduling decisions in each time unit for matching deployed APs and MUs, where the APs are eventually connected to medical platforms.

Ø The proposed medical platform makes the scheduling decision according to the principle of max-weight, which considers data rates between APs and MUs and the queuebacklog size for medical devices.

Advantages

Ø Efficient medical platform design

Ø Address the problem of buffer management and scheduler design

Ø Avoid data overflow and loss in medical devices.

HARDWARE REQUIREMENTS

Processor                     : Any Processor above 500 MHz.

Ram                              :  128Mb.

Hard Disk                    :  10 Gb.

Compact Disk             :  650 Mb.

Input device                :  Standard Keyboard and Mouse.

Output device             :  VGA and High Resolution Monitor.

SOFTWARE SPECIFICATION

Operating System            : Windows Family.

Programming Language   : JDK 1.5 or higher

Database                  : MySQL 5.0