10%

Try : Insurtech, Application Development

Edtech(5)

Events(31)

Interviews(10)

Life@mantra(10)

Logistics(1)

Strategy(14)

Testing(7)

Android(43)

Backend(28)

Dev Ops(2)

Enterprise Solution(20)

Frontend(28)

iOS(39)

Javascript(13)

AI in Insurance(26)

Insurtech(57)

Product Innovation(34)

Solutions(13)

Augmented Reality(7)

Customer Journey(7)

Design(6)

User Experience(21)

Artificial Intelligence(95)

Bitcoin(7)

Blockchain(14)

Cognitive Computing(7)

Computer Vision(6)

Data Science(14)

FinTech(41)

Intelligent Automation(25)

Machine Learning(43)

Natural Language Processing(10)

E-health(2)

HealthTech(5)

mHealth(3)

Telehealth Care(1)

Key Takeaways from the Webinar – Digital Health Beyond COVID-19: Bringing the Hospital to the Customer

8 minutes, 19 seconds read

The outbreak of pandemic COVID-19 put the healthcare sector into a tizzy. From dealing with a rising number of patients and lack of medical supplies and infrastructure to falling in-patient and out-patient footfalls and elective surgeries; it was indeed a difficult period. The latter affected the business of private healthcare due to their inability to sustain cash flow and fixed costs. However, the healthcare system did come around and in response to the COVID-19, new care packages were introduced which included services such as Medical Kits, App to Check Vitals, Home Delivery of Medicines, Remote Doctor Consultations, Remote Nurse Consultations, Helpline for Query Resolution, etc. to give a meaningful digital health experience.

Post-COVID: Future of Indian Digital Healthcare

The pandemic has been around for quite some time now and we’re halfway through the New Normal. In India, the consumer-led healthcare ecosystem is changing to consumer-led digital health. At-home/doorstep delivery is reducing the number of physical touchpoints. The rapid growth of telemedicine and preventive healthcare apps are some of the driving forces for the Indian Healthtech market which is expected to grow up to $21bn by 2025 (which is only 3.3% of the total addressable healthcare market of $638 Bn by 2025). One of the untapped opportunities that the healthcare ecosystem could look into is providing out-patient insurance for day-to-day doctor visits or health needs.

Key takeaways from the webinar

Here are some takeaways from a very insightful and interactive webinar on Digital Health-

Evolving digital health behavior

During Pandemic: 35% of Consumers are impacted at their jobs, 25% are still saving & stockpiling essentials, 38% are in a ‘hibernating & spending’ state, while the remaining 2% are least affected and will continue their old spending behaviors. Here’s a brief overview of changing patient behavior and benefits of patient-generated data-

Patterns of Consumption and Health-seeking Behaviour

People have certainly become more health-conscious since the outbreak of the pandemic. A major shift in consumer behavior is that now they are data-oriented. Tracking health conditions and using medical records to make decisions via mobile apps is trending. Consumers want ownership of their health data. Amongst at-home consumers (in the last three months) 

  • at least 1 in 3 have used a fitness app
  • at least 1 in 4 have used a telehealth service
  • at least 1 in 5 have consulted with a virtual doctor

Doctors’ digital behavior is evolving too

In India, doctors’ are now finding clinical information and engaging with medical reps digitally. Each general physician on their platform now consults around 100 patients. Doctors too are leaning towards digital health platforms to increase engagement with new prescription influencers and auxiliary stakeholders such as pharmacists (formularies) and insurers. Telehealth apps and e-pharmacies will see a boom in the upcoming months as both patients and doctors are getting accustomed to virtual health trends.

Rise of mHealth Apps

mHealth market in India is expected to touch $2.4B by 2024. mHealth apps are more popular in preventive healthcare space. There are 400,000 mobile health apps in India, for self-monitoring a variety of health data — heart rate, bp, sleep pattern, blood glucose level, etc. Some of the top use-cases of mhealth apps are in the fitness and nutrition areas. Now there’s increased adoption of apps in mental health, e-prescriptions, and diagnostics as more consumers demand door-step services. 

Real World Use Cases for Pre and Post-Op Care

Even though the economy has slowly started to open up, the COVID scare has not yet gone. People are still being cautious about stepping out of their homes. We have seen the maximum usage of mobile apps and door-step services during this period. Demand for home care services driven over mobile applications is on the rise. This has opened up so many opportunities in the digital health space to making healthcare accessible from anywhere for both patients and doctors.

Let’s take a look some of the use cases in Digital Health-

Using AI for Doctor Consultations

Many doctors face challenges in managing high patient volume. Digital self-care tools such as an AI doctor can assist here by coming up with a basic diagnosis and treatment plan. These results are then validated by an actual doctor. 

E.g. around 20% of consultations on 1mg is done by an AI doctor.

Know how Mantra Labs helped PAHOMIQ build and train AI models to enhance their Image processing techniques to allow earlier detection of abnormalities and treatment monitoring.

Digital Mental Health Therapy Chatbots

Mental Health and Emotional Wellness were not considered serious health conditions. But in the light of recent events, with depression and suicide rates going up, detecting mental health conditions early-on has become very important. An AI conversational chat tool can help users by monitoring their moods coupled with self-care exercises for dealing with mental health issues. Furthermore, AI can help give people preliminary diagnoses that open ups treatment options, freeing up resources for mental health providers. AI can even triage patients based on the data it collects from patients which will enable doctors to come up with treatment plans.

Ex: Wysa — App supports and encourages users to achieve defined mental health goals.

ORCHA, the World’s leading health app evaluation, and advisor organization awarded Wysa an overall rating of 93%, including 100% on clinical safety. Wysa has also rated the best app for COVID-19 stress and anxiety.

Digital Lab

Managing digital pathology workflows is a challenge for both patients and doctors as it requires a lot of coordination and is time-consuming. A mobile app with features of a digital lab such as case-based tracking, schedule appointments, extract diagnostic data from pathology reports, receive alerts, prescription, billing & inventory management will help both patients and doctors save time and effort. 

Know more about how Mantra Labs built Manipal’s Digital 360° Patient Management app which included features such as scheduling appointments and uploading medical records. Thus, improving the quality of care and patient satisfaction. 

Mobile Phlebotomist App

Tier 2 and Tier 3 markets have limited access to healthcare services. One way to make medicine and testing more accessible to people in these markets is by mHealth apps. The appointment booking feature can help a phlebotomist to manage online patients bookings. The phlebotomists will collect blood samples following a specific protocol and use bar-coded stickers with patient identification information that they stick to collection vials. This enables the field team to manage home sample collection requests, view daily collection schedules and drop-off points, with geo-location services.

Mantra Labs Five-Factor Model for Digital Health Mobile App Transformation

Building Truly Engaging Apps

There are many healthcare apps in play today running from fitness and nutrition trackers to hospital apps. But, how many truly keep the user engaged? 

Personalization

A one-size-fits-all approach will not engage with the majority of the population due to diverse personas. With features like providing timely & relevant advice will make users feel privileged keeping them engaged.

Digital Nudges 

One of the reasons why Health and Wellness apps are more popular is because they help users to self-monitor their progress. Apps with gentle reminders or notifications act as triggers to get desired results.

Empathy by design 

A deep study into the user behavior through surveys/secondary research can help get qualitative data which is crucial to incorporate into the app. If the app matches the user needs, then they are more likely to stay on it.

Frictionless Touchpoints

Lengthy workflows, frequent pop-ups, and advertisements can create a distraction for a user while on the app. An app should not interfere with a user’s daily routine or disrupt their behavior. 

Gamification

The gamified approach grabs a user’s attention and keeps him interested in the app. A plain layout will wear out a user soon. Gameplay focuses on the user’s attention better.

E.g. Using reward badges for completing virtual exercises, with an added caveat like, “This badge will expire in 6 days. Keep exercising to earn more badges.”

In a nutshell

As we move towards more digitization of the healthcare sector, it is important to determine specific KPIs for mobile apps and digital platforms. Good KPIs will inform the need to change before a material loss occurs. Digital transformation is a multi-year effort, and the return on investment is not always immediate. Data is going to be the key. Turning data into actionable information through advanced analytics will help in a long-term improvement strategy. Ultimately, setting the right business outcomes and working backward to solve the problem areas through technology will help achieve those goals. 

Know more about our work in Digital Health and how we have helped clients such as Suraksha Diagnostics, Abbvie, Religare Health Insurance, and SBI Health Insurance build mobile and web applications improving their operational efficiency and customer experience.

Check out the webinar on our YouTube channel.

Further Readings:

  1. Building Consumer Trust in the Digital Healthcare Era
  2. HealthTech 101: How are Healthcare Technologies Reinventing Patient Care
  3. Virtual health: Delivering care through technology
  4. How Mobile Micro-Health Insurance can unlock ‘Digital for Bharat’?

Cancel

Knowledge thats worth delivered in your inbox

Tabular Data Extraction from Invoice Documents

5 minutes, 12 seconds read

The task of extracting information from tables is a long-running problem statement in the world of machine learning and image processing. Although the latest accomplishments in the field of deep learning have seen a lot of success, tabular data extraction still remains a challenge due to the vast amount of ways in which tables are represented both visually and structurally. Below are some of the examples: 

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Invoice Documents

Many companies process their bills in the form of invoices which contain tables that hold information about the items along with their prices and quantities. This information is generally required to be stored in databases while these invoices get processed.

Traditionally, this information is required to be hand filled into a database software however, this approach has some drawbacks:

1. The whole process is time consuming.

2. Certain errors might get induced during the data entry process.

3. Extra cost of manual data entry.

 An invoice automation system can be deployed to address these shortcomings. The idea is to upload the invoice document and the system will read and generate the tabular information in the digital format making the whole process faster and more cost-effective for companies.

Fig. 6

Fig. 6 shows a sample invoice that contains some regular invoice details such as Invoice No, Invoice Date, Company details, and two tables holding transaction information. Now, our goal is to extract the information present in the two tables.

Tabular Information

The problem of extracting tables from invoices can be condensed into 2 main subtasks.

1. Table Detection

2. Tabular Structure Extraction.

 What is Table Detection?

 Table Detection is the process of identifying and locating tables that are present in a document, usually an image. There are multiple ways to detect tables in an image. Some of the approaches make use of image processing toolkits like OpenCV while some of the other approaches use statistical models on features extracted from the documents such as Text Position and Text Characteristics. Recently more deep learning approaches have been used to detect tables using trained neural networks similar to the ones used in Object Detection.

What is Table Structure Extraction?

Table Structure Extraction is the process of extracting the tabular information once the boundaries of the table are detected through Table Detection. The information within the rows and columns is then extracted and transferred to the desired format, usually CSV or Excel file.

Table Detection using Faster RCNN

Faster RCNN is a neural network model that comes from the RCNN family. It is the successor of Fast RCNN created by Ross Girshick in 2015. The name Faster RCNN is to signify an improvement over the previous model both in terms of training speed and detection speed. 

To read more about the model framework, one can access the paper Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks.

 There are many other object detection model architectures that are available for use today. Each model comes with certain advantages and disadvantages in terms of prediction accuracy, model parameter size, inference speed, etc.

For the task of detecting tables in invoice documents, we will select the Faster RCNN model with FPN(Feature Pyramid Network) as a feature extraction network. The model is pre-trained on the ImageNet corpus using ResNET 101 architecture. The ImageNet corpus is a public dataset that consists of more than 20,000 image categories of everyday objects.  We will therefore make use of a Pytorch framework to train and test the model.

The above mentioned model gives us a fast inference time and a high Mean Average Precision. It is preferred for cases where a quick real time detection is desired.

First, the model is to be trained using public datasets for Table Detection such as Marmot and UNLV datasets. Next, we further fine-tune the model with our custom labeled dataset. For the purpose of labeling, we will follow the COCO annotation format.

Once trained, the model displayed an accuracy close to 86% on our custom dataset. There are certain scenarios where the model fails to locate the tables such as cases containing watermarks and/or overlapping texts. Tables without borders are also missed in a few instances. However, the model has shown its ability to learn from examples and detect tables in multiple different invoice documents. 

Fig. 7

After running inference on the sample invoice from Fig 6, we can see two table boundaries being detected by the model in Fig 7. The first table gets detected with 100% accuracy and the second table is detected with 99% accuracy.

Table Structure Extraction

Once the boundaries of the table are detected by the model, an OCR (Optical Character Reader) mechanism is used to extract the text within the boundaries. The text is then processed using the information that is part of a unique table.

We were able to extract the correct structure of the table, including its headers and line items using logics derived from the invoices. The difficulty of this process depends on the type of invoice format at hand.

There are multiple challenges that one may encounter while building an algorithm to extract structure. Some of them are:

  1. The span of some table columns may overlap making it difficult to determine the boundaries between columns.
  2. The fonts and sizes present within tables may vary from one table to another. The algorithm should be able to accomodate for this variation.
  3. The tables might get split into two pages and detecting the continuation of a table might be challenging.

Certain deep learning approaches have also been published recently to determine the structure of a table. However, training them on custom datasets still remains a challenge. 

Fig 8

The final result is then stored in a CSV file and can be edited or stored according to one’s convenience as shown in Fig 8 which displays the first table information.

Conclusion

The deep learning approach to extracting information from structured documents is a step in the right direction. With high accuracy and low running time, the systems can only learn to perform better with more data. The recent and upcoming advancements in computer vision approaches have made processes such as invoice automation significantly accessible and robust.

About the author:

Prateek Sethi is a Data Scientist working at Mantra Labs. His work involves leveraging Artificial Intelligence to create data-driven solutions. Apart from his work he takes a keen interest in football and exploring the outdoors.

Further Reading:

Cancel

Knowledge thats worth delivered in your inbox

Loading More Posts ...
Go Top
bot

May i help you?

bot shadow

Our Website is
Best Experienced on
Chrome & Safari

safari icon