Try : Insurtech, Application Development

Edtech(3)

Events(27)

Life@mantra(8)

Logistics(1)

Strategy(6)

Testing(5)

Android(42)

Backend(29)

Dev Ops(2)

Enterprise Solution(13)

Frontend(29)

iOS(37)

Javascript(13)

Augmented Reality(7)

Customer Journey(7)

Design(5)

User Experience(19)

AI in Insurance(19)

Insurtech(47)

Product Innovation(27)

Solutions(7)

Artificial Intelligence(81)

Bitcoin(7)

Blockchain(14)

Cognitive Computing(6)

Computer Vision(5)

Data Science(11)

FinTech(38)

Intelligent Automation(19)

Machine Learning(42)

Natural Language Processing(5)

Modern Medical Enterprises Absolutely Need Test Automation. Here’s Why.

Nivin Simon
3 minutes, 38 seconds read

The healthcare industry is getting a comprehensive digital facelift. Digital Health Systems (DHS) that use new digital technologies like artificial intelligence & robotics are delivering smarter healthcare services and better health outcomes to the masses. Health organizations are increasingly relying on them to improve care coordination, chronic disease management and the overall patient experience. These health systems are also alleviating repetitive administrative tasks from the roles of healthcare professionals, allowing them more time to practice actual healthcare.

The Modern Medical Enterprise draws on digital-enabled technologies such as telemedicine, AR/VR and remote-monitoring wearables to diagnose diseases and promote self-care. These applications rely on high-volume processing of patient data on a frequent basis.  Healthcare organizations also need to share/receive this information securely over a distributed network. However, sharing patient information remains a challenge, while the inability to access these records in a time-sensitive manner can affect the time-to-treatment for patients.

Deploying digital health systems that are both compliant to regulatory standards and functionally stable for a large number of concurrent users requires significant manned effort. Moreover, QA teams comprised of manual testers may end up working on repetitive manual test case scenarios that can lead to challenges in scaling or rolling out new features. 

How can the modern healthcare enterprise keep pace with issues posed by the safe deployment of their digital health systems? Automated Testing is a hallmark process of any digital transformation project. It gives enterprises the ability to shorten their release cycles and meet their business needs without affecting productivity or operations across the healthcare value chain. Test Automation also allows medical enterprises to run repeatable and extensible test cases against real-world scenarios.

Test Automation Use Case

The growth of DevOps and the rise of mobile-first applications are responsible for driving the growth of the test automation market globally. Today, enterprises are able to go faster-to-market owing to the technological advancements in quality assurance & testing.

For instance, in the case of a large US-based teleradiology firm that offers enterprise Imaging Solutions for improving patient care — a stable and reliable system mandated custom-built test automation frameworks. The medical technology company provides fast & secure access to diagnostic quality images using any web enabled device. To achieve this, they have built a cloud-based image sharing platform that allows digital image streaming, diagnostic & clinical viewing, and archiving for healthcare organizations.

Medical Image sharing among healthcare organizations is altogether brimming with security risks, and requires a complex network of systems to facilitate its smooth functioning. 

medical imaging system architecture
Medical Image Sharing Process among Healthcare Organizations

Also read – How are Medical Images shared among Healthcare Enterprises? 

In order to fulfil their business objectives, Mantra Labs identified key challenges for their testing requirements, namely —

1. Scalability
The platform must be able to support a high number of concurrent users.

2. Fail-over Control

The platform should behave functionally correct under very high loads with stable fail-over capability.


3. Efficiency & Reliability
The platform must scale rapidly when supporting a large user base & multiple formats with minimal page navigation response time.

Several testing components were deployed along with test automation techniques to address the full range of QA issues, including: functional testing, integration testing, GUI testing, and regression testing. 

Mantra Labs created a federated architecture to ensure near-perfect scaling, and true load & data isolation between different tenant organizations. The federated architecture consists of a number of deployments and a central set of components that stores global information like lists of organizations & users, and provides a centralized messaging service. 

test automation process flow diagram for modern medical enterprises
Mantra Labs Test Automation Process

Test Automation Improves Accuracy & Test Coverage

The entire cycle of bug detection in the UI, API and Server Loads involves several weeks of regression manual efforts. By automating tests, techniques like Stochastic Tests can be applied to detect bugs and reduce the overall cycle time.

Through Mantra Labs deep medical domain expertise, in-depth testing practices, intuitive suggestions for platform scaling and successful test automation efforts — significant business objectives were realised over the course for the client. Mantra was able to achieve over 60% reduction in cycle time, and about 65 per cent improvement in bug detection capability before the release cycle.

Nearly 35% of Executive Management objectives revolve around implementing quality checks early in the product life cycle, which can be achieved through test automation. For further queries and details about automated testing, please feel free to reach us at hello@mantralabsglobal.com

Cancel

Knowledge thats worth delivered in your inbox

The race for AI dominance is a tough game to lose. For the five biggest giants of the Tech industry at least — Microsoft, Google, Facebook, Amazon, and Apple — it is a race they can’t afford to lose. Each year, they pour billions of dollars into their AI research divisions, each trying to advance the possibility of Artificial General Intelligence (AGI) with the intention to gain first-movers advantage in shaping the business era of the future.

AI has already, to some extent, permeated the modern business landscape. Narrow AI has given the modern consumer unique innovations like Siri and Alexa. Although this version of AI does not really compare with actual human intelligence, what it can do is take a specific skill like facial recognition or voice identification and perform these tasks at a superhuman level. Narrow AI is also more prominently applied across any useful value chain for reducing human effort. By 2021, Gartner predicts that it will save over 6.2 billion hours of worker productivity globally, and create up to $2.9 trillion of business value. 

The investments into advanced AI requires some serious deep pockets to fund these projects. Google’s DeepMind, for instance, burned through $500M in 2018 alone. Most companies would shudder at the thought of burning through that kind of money, but the reality is that for companies who are serious about winning the AI race — half a billion dollars is a small price to pay. 

A lot of companies are also investing heavily into conversational platform capabilities. They believe that as more and more people transit into the mobile-first era, the way people communicate will also change. Microsoft is also leading this charge, through their innovative ‘conversation as a platform’ offering, which anticipates a momentous shift from the app-focused world we are so used to living in now. 

For the Big Five, the race is to create the most advanced brain. A brain that can either match or surpass human intelligence. (Not so subtly, Google’s AI research lab that acquired DeepMind is called Google Brain). This ‘brain’ will live in the cloud and that is where all of this phenomenal effort points to. A connected sphere of trillions of data points being captured and stored on the cloud, where AI systems will comb through the data in near real-time to perform extremely complex tasks in a few seconds. A cautionary note: Even the most sanguine experts opine that AI is still decades away from achieving this level of intelligence.

The AI battlefield is filled with cases of organizations & business leaders who were not bold enough or didn’t possess the adequate vision for competing in the AI space. When Satya Nadella, Microsoft’s current CEO, took over from Steve Ballmer – he corrected Ballmer’s 14 year-long misguided turn that ultimately led to Microsoft failing in the mobile devices space, on his very first day. He made Microsoft reposition itself as cloud and mobile-first. Nadella steered the company back from a $500M company in 2014, to slightly over a trillion dollar valuation in 2020. He spurred a new vision for Microsoft, and it’s future bets — starting with Artificial Intelligence.

However, to be a leader in the AI space, you need the best talent that’s out there working for you and not for the competition. And so, a new race was on — who could acquire and afford the best minds in AI?

The Godfathers of AI

Each year, the Association for Computing Machinery (ACM) presents the Turing award to individual(s) who have made significant contributions to the field of computing. The award is widely considered the Nobel prize of Computing. In 2018, three researchers — Geoffrey Hinton, Yann LeCun and Yoshua Bengio were recognized for their contributions to neural nets and deep learning (a sub- field within AI). 

The trio fostered several breakthroughs in the early 90s in areas like computer vision and speech recognition. It is the result of this work that has eventually found its way into almost every smartphone such as facial unlocking features in most devices, and futuristic AI technologies that are still in early development like self-driving cars. Widely respected and deemed within the computing community, the three are often regarded as the ‘Godfathers of AI’. 


Geoffrey Hinton

Geoffrey Hinton, a professor at the University of Toronto, is often recognized as the ‘Godfather of Deep Learning’. In 2012, he and some fellow students entered the ImageNet challenge to prove neural networks were able to achieve more accurate results for object recognition tasks. 

They beat the second best algorithm at the competition by more than 40%. Once the story broke and the World took notice of their work, the techniques pioneered by Hinton, LeCun and Bengio have become the foundational basis of convolutional neural networks and AI in general.

2013-05-23-my+photos+of+flowers

Google took notice, and quickly realised that these algorithms were far superior than what they previously had, and realised they could use it for making their ‘photo searching’ more precise and faster. Hinton’s research, that he himself spun off into a company called DNN Research, would be acquired by Google in 2013, and with that the ability to  pursue true computer vision for their image search features, and get machines to look at the world as humans do.

While LeCun, would eventually join Facebook as it’s Chief AI Scientist, Yoshua Bengio decided to stay neutral and not get involved with corporations. Satya Nadella however pursued Bengio for a few years, before Bengio’s skepticism, although not unprecedented, waned. If companies were truly pursuing AGI, then such technology should not be in the hands of one or two companies alone; that could be disastrous. 

Microsoft makes bold plans for AI

With Bengio at the helm of Microsoft’s AI research, Nadella infused AI within all three core offerings of the company: Office, Windows and Azure. Today, Microsoft’s approach has allowed their products to remain competitive alongside Google ’s search engine, productivity suite tools and personal assistant capabilities. Microsoft leads the AI patent race with around 18,300 patents.


Yoshua Bengio

Microsoft recently demonstrated how Cortana can sit in on meetings, transcribe the entire conversation, add reminders, tasks, and update the meeting attendees with a copy of the minutes – irrespective of language. 

The cloud-based AI era of computing is fast upon us. However, the problem with having AI in the cloud is that latency (time it takes to receive information over the internet) becomes a real problem. With 5G, things can go much faster, but the key lies in it’s tandem with super computing processors. 

Enter Nvidia Volta, which some say is the key to unlocking real AI computing speeds. In fact, the fastest supercomputers on the planet also run on these graphic processors. 

At a crucial turning point in the race, Elon Musk decided to back one of one of Hinton’s former students who was part of the three member team at ImageNet, Ilya Sutskever. Sutskever, even at the time was somewhat of a prodigy in the field, and shared Musk’s concerns of AGI if it fell into the wrong hands. Together, they started Open AI, with a vision to build safe AGI. In 2015, Microsoft invested in the company with deal specifics implying Microsoft could commercialize any Artificial General Intelligence technology that Open AI creates. 

Apple is betting on AI at the edge
Three years ago, Apple was falling behind Amazon, Facebook and Google in terms of AI investments and the failure of their flagship smartphone, the iPhone X, to reveal any new significant technological improvements. To catch up they have spent upwards of $200M to stay in the race. However, Apple, unlike its competitors, is not concerned with moving workloads onto the cloud. While everyone else is betting on the cloud, Apple is betting on hardware. Their AI strategy is to focus on devices and have Machine Learning workloads run locally on these devices. Apple believes this way they won’t compromise on user privacy. To facilitate this, they have launched an exclusive platform called ML Create (to train ML Models) and Core ML (to build AI models into Apple apps).

They have reinforced this strategy with several AI startup acquisitions like Turi and Spectral Edge, including their recent acquisition of Xnor.ai — a startup that builds low power ML hardware and software. Apple has acquired over 20 companies since 2010, to fulfill this vision.

However, with other open source platforms like Google’s TensorFlow, Apple has failed to attract the attention of the AI developer community to latch onto their new ecosystem. TensorFlow came out three years before Apple launched ML Create, and in that time developers have already built a strong community committed to building products on other ecosystems that are not as notoriously closed off as Apple’s devices.
It is too early to tell, if Apple will be able to recover back some of that lost ground, or if the competitive advantage being forged by any of the Big Five at this stage will work out in the near future. The bets are placed, and therace is still on.

Cancel

Knowledge thats worth delivered in your inbox

The need for and consequently the number of solutions for reading hand-written forms in an automated manner has been on a rise for as long as one could remember. Almost all businesses to varying degrees utilize paper-based forms that are filled by customers by hand. Most if not all of these businesses convert this handwritten information into the digital format. Depending on the technological sophistication or the size of the business this digitization might be done manually by one or more data entry specialists or through an automated solution. 

It’s easy to see how the manual route may not be an ideal solution for medium or large-sized business. Some of the apparent drawbacks of manual document processing are:

  1. The cost of having data entry specialists quickly add up as more documents need to be digitized necessitating adding more resources.
  2. Manual data entry is a slow process.
  3. Manual data entry is error-prone and requires a quality inspection which is costly and not fail-proof.

Many businesses have realized this and have transitioned to some form of a partially or fully automated solution to this problem. However, it’s not all rosy for these businesses either. The problems these businesses face is primarily related to the accuracy of the current solutions in the market. 

Shortcomings of Existing Hand-written Document Processing Solutions

The industry average for ICR (Intelligent Character Recognition) accuracy at the character level is about 70% and it will drop significantly if measured at word level which is what matters at the end. Such automation may allow for reducing the number of data entry personnel but with such a low level of accuracy, there will be a need for increased quality check resources, which are often more expensive than data entry resources hence diluting the cost-benefit of automation. Moreover, since the quality check is a slower process than data entry, this kind of automation doesn’t even address the speed problem.

Some of the reasons that result in a low level of accuracy among existing document processing solutions are:

  • Poor form design
  • User input not in line with the format
  • Noisy images
  • Misaligned documents
  • Low-quality scanning of documents
  • Spelling mistakes by the user
  • Overwriting/corrections by user

While we may not have control over some of the above factors such as form design and user input, we can definitely improvise the data extraction models to account for the other factors such as image noise, misalignments, spelling mistakes etc.

Our ICR Solution

The Document Parser solution in FlowMagic provides an intuitive user interface where data can be extracted from any standard form in three easy steps:

Step 1:   The user annotates the form (this is a one-time exercise for each new form) using an easy and intuitive UI. During annotation, each input field can optionally be labelled as mandatory. The user can specify the datatype for each field as alphabets, numeric or checkbox and also set the context for the field e.g. Name, PAN, City, Car Make, Date etc. Once done, the saved template can be used repeatedly for reading forms of the same type as long as there are no changes in the form design. In case of a change, the saved template can be easily modified. 

Step 2:   The user uploads one or more forms and chooses the corresponding template (from previous annotations). The system automatically extracts data from the forms.

Step 3:  The system exports the output in CSV, XML or JSON as desired by the user. If any field was marked as mandatory during annotation, the system also outputs a list of all mandatory fields that are blank.

Salient features of ICR Document Parser

  1. The standard form being annotated can be any number of pages. The input form need not have the same number of pages. If there is a mismatch between the pages in the input form and the template, the system does a matching and runs the data extraction on matching pages only. This also means that the input form need not be sorted correctly.
  2. The system can read handwritten as well as printed forms.
  3. The system corrects for minor misalignments during scanning of documents or documents scanned in the wrong orientation.
  4. The system has inbuilt dictionaries for various contexts such as Name, Cities, States, Countries, PAN, Profession, Marital Status, Relationship, Amount, Car Make, Date, Gender.
  5. The various data types supported by the system are alphabets, numeric, alphanumeric, checkboxes and special characters.
  6. The system corrects user errors or scanning issues by performing data type and dictionary checks (see examples below).
  7. The system checks for mandatory fields to make sure the form is completely filled.

Examples of Data Read/Corrections Made by an ICR

Benefits of ICR

Flexibility – you can annotate a wide variety of forms with complex inputs and data formats using the multiple data types and contexts built into the system.

Speed – Both annotation and data extraction are very user-friendly and fast. The system can extract data from a five-page form in under 30 seconds.

Scalability – The system is highly extensible and once set up for one type of form can easily be scaled for multiple forms or to process documents in bulk of the same format.

Accuracy – The character level accuracy of our model is over 90%. Word level accuracy depends on the form design and quality but in general, varies between 75% and 85%.

Workflow

ICR (Intelligent Character Recognizer) workflow

No matter what solution you use, you can always benefit from these best practices for form design to improve the accuracy of your ICR:

  1. Have all instructions in bold at the top of the form.
  2. Instruct the user to write clearly in block letters as the form will be processed by a machine.
  3. Provide examples of how to enter data wherever there is a scope for confusion.
  4. Instead of providing a free form space for data entry, it provides a clearly marked space with a specific location to enter each character.
  5. The overall space should be large enough to contain the requisite data to avoid user writing outside of this space.
  6. Have enough separation between the space for two fields to avoid overlap.

To learn more about how FlowMagic can improve the accuracy and speed of your document digitization/Intelligent Character Recognition (ICR) or discuss your broader AI goals, please get in touch with us 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