Try : Insurtech, Application Development

Edtech(3)

Events(25)

Life@mantra(7)

Logistics(1)

Strategy(2)

Testing(4)

Android(42)

Backend(26)

Dev Ops(2)

Enterprise Solution(12)

Frontend(29)

iOS(36)

Javascript(12)

Augmented Reality(7)

Customer Journey(6)

Design(5)

User Experience(19)

AI in Insurance(17)

Insurtech(46)

Product Innovation(26)

Solutions(5)

Artificial Intelligence(76)

Bitcoin(7)

Blockchain(14)

Cognitive Computing(6)

Computer Vision(5)

Data Science(11)

FinTech(36)

Intelligent Automation(19)

Machine Learning(41)

Natural Language Processing(5)

How are Medical Images shared among Healthcare Enterprises

For the modern healthcare organisation, extending better patient care across the service continuum involves — new challenges that surround sharing information over a distributed network. Effectively sharing patient information remains a challenge, while an inability to access these records in a time-sensitive manner results in re-imaging and re-testing the patients — affecting both ‘time-to-treatment’ and the bottom line.

The release process for medical images is altogether complicated — brimming with security related-risks — as images (such as X-Ray Scans, MRI scans, PET scans, etc.), are created and released across several departments and systems, while being purposefully kept ‘out-of-reach’ from a host of unauthorised users.

Training & controls on release policies and procedures require ‘health information management’ expertise — since, Image Handling (electronically) can become susceptible to data corruption, complex accessibility/sharing issues and high security risks —  all of which raises potential red flags for health information management (HIM) professionals.

So how does Medical Image sharing work in this environment, and what, if any — are the safeguards surrounding its ‘release’ process?

First, let’s delve into the term ‘Medical Imaging’ itself. According to the WHO, the technique embodies different imaging modalities and processes to image the human body (creating visual representations) for diagnostic and treatment purposes — making it crucial for improving public health initiatives across all population groups.

Once the image has been captured using a medical imaging device (routine imaging techniques like ultrasound, MRI etc.), then it is necessary to archive and store the images for future use and further processing. Unlike regular images (that we’re more familiar with) stored in more conventional formats like jpeg, png, etc. — medical images are stored in a format known as ‘DICOM’ (Digital Imaging and Communication in Medicine) standard. The medical practitioner responsible for acquiring and interpreting such medical images is a ‘Radiologist’ — while the system they rely on for electronic image data storage is called ‘PACS’ (Picture Archiving and Communication System).

In other words, if a healthcare organization or an outside consultant (physician, clinician) needs access to an individual patient’s medical images, then the access and retrieval will have to go through PACS (typically controlled and operated by the Radiologist itself).

Here is a simple process diagram of a medical imaging system —


A Typical HIPAA-compliant Medical Imaging Management System places a request (for a specific file) to ‘PACS’ via an intermediary system known as ‘Edge Server’. The sole purpose of the Edge Server is to function as a request node, so that other hospitals or physicians can contact the particular radiologist (who possesses the images stored in PACS) and place a request to access a copy of the file in question.

Critical use cases arise for medical image sharing involving support for:

  • Remote image viewing (out of network)
  • Specialist consults
  • Telehealth (examples such as teleburn, telestroke)
  • Trauma transfers
  • Ambulatory image review

Each of the three highlighted sections (see diagram) can perform various functions, while communication is defined through specific rules and standards that are legally enforced and universally followed.

Medical images (digital) are typically stored locally on a PACS for retrieval. A PACS consists of four major components: The imaging modalities such as X-ray plain film (PF), CT and MRI; a secure network for the transmission of patient information; workstations for interpreting and reviewing images; and archives for the storage and retrieval of images and reports. To communicate with the PACS server we use DICOM messages that are similar to DICOM image ‘headers”, but with different attributes.
The Edge Server manages several functions that allow users to sort through hundreds of thousands of large-volume data and retrieve a specific file from a database either stored in ‘PACS’ or on the ‘MIMS’.

Through the ‘Edge Server’, we can access images stored in PACS. The ‘Management Services’ operation is the first and foremost feature — meaning, a user can control & maintain the complete functionality of the server through this. Using ‘Remote Authentication’ —  a networking protocol operating by way of specific ports, users can obtain centralized authorization and authentication to request files from PACS.

To verify basic DICOM connectivity to the server — i.e, to check if the server is live or not, a C-Echo message is sent to ping the server, after which it will wait for its response. Once the server is identified as live, a user can perform querying and retrieval based operations. Next, the user can begin the process of requesting DICOM images from the Medical Image Management System — known as ‘Ingestion’. DICOM Ingestion is performed via pre-assigned IP and port addresses (default ports are 2104-2111).

Basic DICOM Operations

To check if the specific image or set of images is actually located on the particular server – a query-based C-FIND operation is performed by sending a request to the server. First, the user establishes a network connection to the PACS server, following which a C-FIND request message (which is a list of DICOM attributes) is prepared. The user fills in the C-FIND request message with ‘keys’ that match. (for e.g. to query for a patient ID, the patient ID attribute is filled with the patient’s ID). Then, the C-FIND request message is sent to the server.

The server sends back to the user, a list of C-FIND response messages, each of which is also a list of DICOM attributes, populated with values for each match. The Images are then retrieved from the PACS server through a C-MOVE request, using the DICOM network protocol. Retrieval can be performed at the Study, Series or Image (instance) level. The C-MOVE request specifies where the retrieved instances should be sent (using separate C-STORE messages). The C-STORE operation, also known as DICOM Push is used to simply push (send) the images to the PACS server (or P2P — Push to PACS). The DICOM storage service is implemented through the C-STORE message — the SCU sends a C-STORE-RQ (request) message to the server, which includes the actual dataset to transfer, and the server answers returning a C-STORE-RSP (response) message to the user, communicating success or failure of the storage request.

These DICOM functions allow health management professionals, physicians, radiologists and indirectly their patients to benefit from utilising secure protocols in handling confidential medical image data — extending the ability to view such images discreetly and instantly; avoiding duplication costs; and reducing unnecessary radiation exposure to patients.

Medical Image Sharing furthers the “Health 2.0” initiative by being able to instantly and electronically exchange medical information between physicians, as well as with patients — improving communication within the industry.

Cancel

Knowledge thats worth delivered in your inbox

Chatbots are the assistants of the future and they are taking the Internet by storm. Ever since their first appearance in 1994, the goal was to create an AI that could conduct a real dialogue with their interlocutors. The purpose is to free up customer service agents’ time so they could focus on more delicate tasks- which require a more human approach.

If you are thinking about including a chatbot on your website, here are the things you need to keep in mind to boost customer engagement and deliver high-quality services.

Define your audience

First things first- think about who will be interacting with the chatbot? Who are your customers? How do they talk? How can you address them in a way they’ll enjoy? How can you help them?

For instance, if your company sells clothes that are mostly designed for young adults, using a less formal tone will be much more appealing to them.

Lisa Wright, a customer service specialist at Trust My Paper advice: “Customer service calls are usually recorded, so listening to a few of them can be a good place to start designing your chatbot’s lines of dialogue.”

Give your bot some character

People don’t like to talk to plain, simple robots. Therefore, giving your chatbot some personality is a must. Some brands prefer naming their chatbots and even design an animated character for them. This makes the interaction more real.

For example, The SmarterChild chatbot- designed back in 2000, was able to speak to around 2,50,000 humans every day with funny, sad, and sarcastic emotions.

However, the chatbot’s character needs to match your brand identity and at the same time- appeal to customers. Think about – how would the bot speak, if they were real? Are there some phrases or words they would never use? Do they tell jokes? All these need to be well-thought through, before going into the chatbot writing and design phase.

According to a report published by Ubisend in 2017, 69% of customers use the chatbot to get an instant answer. Only 15% of them would interact for fun. Thus, don’t sacrifice the performance for personality. 

Also read – 5 Key Success Metrics for Chatbots

Revise your goals before chatbot writing

Alexa- Amazon bot has 30+ skills which include scheduling an appointment, booking a cab, reading news, playing music, controlling a smartphone, and more. However, every business bot doesn’t need to be a pro in every assisting job.

Before entering the writing phase, think over once again – WHY you need a chatbot? Will it help customer service only? Or will it also help in website navigation, purchase, return, refund, etc.?

Usually, customers want one of the three things when they visit your site: an answer to something they’re looking for, make a purchase, or a solution to their problem. You can custom build your chatbot to tackle either one or all of these three situations. Many brands use chatbots to create tailored products for their clients.  

Cover all possible scenarios

When you start writing the dialogue, consider the fact that a conversation can go in many directions. To ensure that all the situations are covered- start with a flowchart of all possible questions and the answers you chatbot can give.

To further simplify your chatbot writing, take care of one scenario at a time and focus on keeping the conversation short and simple. If the customer is too specific or is not satisfied with the bot’s response, do not hesitate to redirect them to your customer service representatives.

For instance, Xiaocle is one of the most successful interactive chatbots launched by Microsoft in July 2014. Within three months of its launch, Xiaocle accomplished over 0.5 billion conversations. In fact, speakers couldn’t understand that they’re talking to a bot for 10 minutes.

Also read – Why should businesses consider chatbots?

This article is contributed to Mantra Labs by Dorian Martin. Dorian is an established blogger and content writer for business, career, education, marketing, academics, and more.

Cancel

Knowledge thats worth delivered in your inbox

The antiquated commodity of Financial ‘Coverage & Protection’ is getting a new make-over.  Conventional epigrams like ‘Insurance is sold and not bought’ are becoming passé. Customers are now more open than ever before to buying insurance as opposed to being sold by an agent.  The industry itself is witnessing an accelerated digitalization momentum on the backs of 4G, Augmented Reality, and Artificial Intelligence-based technologies like Machine Learning & NLP.

As new technologies and consumer habits keep evolving, so are insurance business models. The reality for many insurance carriers is that they still don’t understand their customers with great accuracy and detail, which is where intermediaries like agents and distributors still hold incredible market power.

On the other hand, distribution channels are turning hybrid, which is forcing carriers to be proficient in their entire channel mix. Customer expectations for 2020 will begin to reflect more simplicity and transparency in their mobility & speed of service delivery.

A recently published Gartner Hype Cycle highlights 29 new and emerging technologies that are bound for greater business impact, that will ultimately dissolve into the fabric of Insurance.

For 2020 and beyond, newer technologies are emerging along with older but more progressively maturing ones creating a wider stream of opportunities for businesses.

Gartner-Hype-Cycle

Irrespective of the technology application adopted by insurers — real, actionable insights is the name of the game. Without it, there can be no long term gains. Forrester research explains “Those that are truly insights-driven businesses will steal $1.2 trillion per annum from their less-informed peers by 2020”.

Based on the major trends identified in the Hype Cycle, 5 of the most near-term disruptive technologies and their use cases, are profiled below.

  1. Emotion AI
    Emotion Artificial Intelligence (AI) is purported to detect insurance fraud based on the audio analysis of the caller. This means that an AI system can decisively measure, understand, simulate and react to human emotions in a natural way.

    F0r Insurers, sentiment and tone analysis captured from chatbots fitted with emotional intelligence can reveal deeper insights into the buying propensity of an individual while also understanding the reasons influencing that decision.

Emotion-Intelligence-Market



Autonomous cars can also sensors, cameras or mics that relay information over the cloud that can be translated into insights concerning the emotional state of the driver, the driving experience of the other passengers, and even the safety level within the vehicle.

Gartner estimates that at least 10% of personal devices will have emotion AI capabilities, either on-device or via the cloud by 2022. Devices with emotion AI capacity is currently around 1%.

  1. Augmented Intelligence
    Augmented Intelligence is all about process intelligence. Widely touted as the ‘future of decision-making’, this technology involves a blend of data, analytics and AI working in parallel with human judgement. If Scripting is rules based automation, then ‘Augmenting’ is engagement and decision oriented.

    This manifests today for most insurance carriers as an automated back-office task, but over the next few years, this technology will be found in almost all internal and customer facing operations. Insurers can potentially offer personalised services based on the client’s individual capacity and exposure to risk — creating opportunities for cross/up-selling.
Gartner-Data-Analytics-Trends-Forecast-2019


Source: Gartner Data Analytics Trends for 2019


For instance, Online Identity Verification is an example of a real-time application that not only enhances human’s decision making ability, but also requires human intervention in only highly critical cases. The Global value from Augmented AI Tools will touch $4 Trillion by 2022.

  1. AR Cloud
    The AR Cloud is simply put a real-time 3D map of an environment, overlayed onto the real World. Through this, experiences and information can be shared without being tied down to a specific location. Placing virtual content using real world coordinates with associated meta-data can be instantly shared and accessed from any device.

    For insurers, there is a wide range of opportunities to entice shopping customers on an AR-Cloud based platform by presenting personalized insurance products relevant to the items they are considering buying.

    The AR ecosystem will be a great way to explain insurance plans to customers, provide training and guidance for employees, assist in real-time damage estimation, improve the quality of ‘moment-of-truth’ engagements. This affords modern insurance products to co-exist seamlessly along the buying journey.

  2. Personification
    Personification is a technology that is wholly dependent on speech and interaction. Through this, people can anthropomorphize themselves and create avatars that can form complex relationships. The Virtual Reality-based concept will be the next way of communicating and forming new interactions.

    VR Applications such as  accident recreation, customer education and live risk assessment, can help insurers lower costs for its customers and personalise the experience.

    Brands have already begun working their way into this space, because as they see it — if younger generations are going to invariably use this technology for longer portions of their day for work, productivity, research, entertainment, even role-playing games, they will shop and buy this way too.

  3. Flying Autonomous Vehicles and Light Cargo Drones
    Although this technology is only a decade away from being commercially realized, the non-flying form is about to make its greatest impact since its original conception. Regulations are the biggest obstacle to the technology taking off, while its functionality continues to improve.

    The Transportation & Logistics ecosystem is on the brink of a complete shift, which will create a demand for a wide array of insurance related products and services that covers autonomous vehicles and cargo delivery using light drones.

While automation continues to bridge the gaps, InsurTechs and Insurance Carriers will need to embrace ahead of the curve and adopt newer strategies to drive sustainable growth.

Mantra Labs is an InsurTech100 company solving complex front & back-office processes for the Digital Insurer. To know more about our products & solutions, drop us a line at hello@mantralabsglobal.com

Cancel

Knowledge thats worth delivered in your inbox

Loading More Posts ...
Go Top

May i help you?

Our Website is
Best Experienced on
Chrome & Safari

safari icon