ICACT20210192 Slide.18
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Thanks for your attention! Now! Floor is open for QnA. Welcome any comment or question.
|
|
ICACT20210192 Slide.17
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
In this paper, a solution based on blockchain and AI Federated learning has been proposed to protect and share safely the patient’s healthcare data.
After analysing the potentials of combining blockchain and AI, we realized that the proposed solution would help to build robust models and share them without being compromised. However, the application wasn’t developed and tested, yet this solution is indeed promising.
In the future, our proposed framework will be implemented and tested in a real-world environment and its effectiveness will be studied more in detail.
|
|
ICACT20210192 Slide.16
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Recently NVIDIA Clara FL – A FL based framework has been using for testing purposes in medical devices. If it is successful then the toolkits developed using this framework will help the radiologist to develop the algorithm in their own institutions using their own patient data.The start-up Owkin works on machine learning developed a federated learning platform that gives the owner the power to track of their data usage.
|
|
ICACT20210192 Slide.15
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
As the AI based model is getting traction these days especially in the field of medical imaging, protecting the private information is necessary before used in the real time environment.Particularly in the time of pandemic, AI based model has been used for detecting Covid-19 using medical imaging, however these models have not been used because of the leakage of the personal information as well as the models are not robust.With the availability of Federated learning it is ensure that robust model could be made without leaking the private information and at the same time combination with the blockchain helps to transact data securely with different service providers.
|
|
ICACT20210192 Slide.14
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
The figure shows the token usage flow.The details are as follows:1) The buyer or user logs in the platform using their credentials.2) The buyer requests the data using their provided token. 3)The ML engine requests the usage rights to the ledger before proceeding.4) If the buyer acknowledged by the system, the blockchain replies with the hash of the requested data.5) The engine queries the data using the hash and other information returned by the blockchain.6)The buyer can download the data or use the ML cloud engine to train it.
|
|
ICACT20210192 Slide.13
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
The overall architecture of the proposed framework shown in figure
|
|
ICACT20210192 Slide.12
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Artificial Intelligence (AI) has the capability to provide radiologists with tools to help improve their productivity and decision making, possibly leading to quicker diagnosis and improved patient outcomes.Benifits are mentioned as follows:
1.Helping to improve productivity of clinical workflows utilizing imaging
2.Assisting in lowering the risk of “negative” clinical consequences associated with delays in radiologist reading, interpreting and reporting
3.Empowering care teams to easily view radiology work product, accelerate clinical decision making, and streamline workflows, helping to result in an improved patient experience and outcomes
|
|
ICACT20210192 Slide.11
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
The architecture for integration of AI to the radiology workflow is shown in the figure
|
|
ICACT20210192 Slide.10
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
The architecture for the radiology workflow is shown in the figure
|
|
ICACT20210192 Slide.09
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
The artificial intelligence(AI)-based technique already providing promising result in detecting various critical disease using medical images.Mainly the role of the AI based technique in diagnosis of various diseases using medical images is to act as a decision support system for the doctors so that the decision taking process will be faster and at the same time maintaining the precision and accuracy.
|
|
ICACT20210192 Slide.08
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
.jpg) |
The architecture for the blockchain utilization in various healthacre applications and architecture for smart contract is shown in the fig 1 and fig 2 respectively.
|
|
ICACT20210192 Slide.07
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Data breaches can cost healthcare organizations $380 per affected record, but current systems are vulnerable to numerous types of attacks.Patient data is extremely valuable to hackers looking for detailed identity information, which makes securing electronic health records (EHRs) and associated personal details a top priority in the healthcare industry.
Blockchain technology possesses features such as decentralized storage, cryptography and smart contracts provide a framework for organizations to improve data protection while maintaining accuracy and preventing unauthorized access to or alteration of patient information
|
|
ICACT20210192 Slide.06
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
The way blockchain works is mentioned as follows:To send or receive bitcoin, a bitcoin user must do a transaction and the user broadcasts it to every part of the Bitcoin Network.For a successful transaction, the user must be added to the public digital ledger i.e. Blockchain. The network of nodes validates the transaction.Transaction data is collected and verified by the use of “Miners” against the existing blockchain ledger and it solves a cryptographic puzzle that allows them to add a recent transaction block to the blockchain.It helps to ensure each sequential block of the public ledger is accurate because of the fact that so many users have the same records and they are many to confirm the same transactions.A new block which is created is added to the existing blockchain as mentioned earlier. Once a block is added to the blockchain it is there forever and becomes a public record.
|
|
ICACT20210192 Slide.05
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
As researchers, we know why we don't have as many images available for training. It's not because we're not collecting enough; although it can be expensive to acquire medical images, healthcare centres already produce millions of scans each year both for direct care and for research worldwide. The images are acquired, they're just not openly available for research (nor should they be). Understandably, access to this data is highly constrained, even within institutions, thanks to important patient data privacy regulations.These privacy concerns are important and necessary, but this has limited us from fully maximizing the benefits of artificial intelligence in our research.
Our solution used federated learning, which allows us to train AI models on distributed datasets that we cannot access directly.
|
|
ICACT20210192 Slide.04
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
In a more general sense, Federated Learning allows for machine learning algorithms to gain experience from a broad range of data sets located at different locations. The approach enables multiple organizations to collaborate on the development of models, but without needing to directly share secure data with each other
|
|
ICACT20210192 Slide.03
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Traditional machine learning involves a data pipeline that uses a central server that hosts the trained model in order to make predictions.The downside of this architecture is that all the data collected by local devices and sensors are sent back to the central server for processing, and subsequently returned back to the devices. This round-trip limits a model’s ability to learn in real-time.Federated learning (FL) in contrast, is an approach that downloads the current model and computes an updated model at the device itself (ala edge computing) using local data.
|
|
ICACT20210192 Slide.02
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Although there is an increase in the amount of data generated from connected devices,it is very challenging to make a robust AI model since most organizations do not want to share their data with other third parties due to privacy concerns.However this this research presents a solution that can protect privacy of the personal healthcare records based on blockchain and Federated Technologies.
|
|
ICACT20210192 Slide.01
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Hello everyone, my name is Satyabrata Aich. For ICACT 2021, I have coauthored with six co-authors.The title of our presentation is "Protecting Personal Healthcare Record Using Blockchain & Federated Learning Technologies.
|
