MEDBOT
AI-powered medical robots on the blockchain
Inspiration
Smart environments of the future for seniors aged 65+. Seniors represent 5% of population yet 42% of the $3.8 trillion healthcare industry. For many seniors who live in low density rural areas, access to doctors is difficult and can takes months. The solution is to build smart spaces for senior care aging in place with a focus on the following:
- Autonomy - they want to stay independent with physical mental and financial freedom
- Safe spaces - reduce environmental risks and homes while supporting mobility
- Long-term health - identify health risks managed multiple medications and chronic illnesses
- Social connection - relationships with friends and family
The solution is Medbot.
Seniors can connect to the internet and integrate with our advanced smart home technology with innovative use of AI-powered medical robots on the robonomics blockchain. One of the main problems with remote medical healthcare is the way highly identifiable sensitive medical patient's data is stored. Instead of records in physical files, we have EPR - an Electronic Patient Record system. The lack of consistent communication between healthcare providers is a large issue in the medical industry.
What it does
Medbot Robonomics is a solution for smart household aging in place environments by using ROS-compatible Arduino IoT controlled by web and mobile interfaces.
Healthcare professionals who access the patient's information through the institution's certified database which can securely offload data from the blockchain. This interface can allow all professionals to have immediate access to patient history and conduct data analysis without changing the data itself. Seniors have a separate interface which allows them to read-only, but have complete access to their own personal history. This increases transparency between medical professionals and patients and allows patients. Information on user friendly web and mobile interfaces to allow for a positive user experience for all stakeholders.
Blockchain technology to store patient profiles within block ledgers to securely store all historical data about a patient like examinations, medical procedures, lab tests, and medications anywhere with the patient's permission. Currently, most IoT and robotics applications are usually organized under a centralized cloud control. Medbot is built on top of Robonomics for a secure way to transfer and centralize patient data to allow access to patient data for all certifies healthcare institutions and patients. Using blockchain encryption for data privacy and point security in order to protect personal identities online. Medical data remains secure and authentic, maintaining data integrity and a chain of trust.
RFID sensors with location tracking for safety with checkpoints. This feature locates missing items with software hardware connectivity IOT - MedBot Robot for home monitoring and medical alerts/notifcations. Serverless IoT architecture includes tagging valuable objects with RFID tags to make them easily locatable this provides Peace of Mind as we're able to support independent living promoting safer more accurate remove healthcare for seniors. Arduino IOT hub integrated with ROS was required to run with audio and visual sensors.
Computer vision with three tensorflow models medical diagnosis and automated pill box and fall detection these features allow those with visual impairments to use their smart devices to navigate around their homes and will enhance the sight seniors already have. Computer Vision with Tensorflow.js with an automated pillbox fall detection, visual impairment aid, and remote medical diagnostics. Natural language processing with Tensorflow JS and Open AI's GPT3 - speech recognition, text to command, voice activation, and a chatbot to combat social isolation. Hardware is also attached to an audio sensor allowing for natural language processing capabilities liketext to command and a chat bot feature with voice activation powered by GPT3.
Blockchain medical data with Robonomics encryption on top of Polkadot and IPFS - allows for data privacy and endpoint security for controlling robotics from distributed clouds. Considerations made for cybersecurity, compliance, connectivity, interoperability, reliability, and single point of failure while leveraging cloud technologies.
How we built it
Refer to technical architecture diagram.
Medblock integates Arduino IOT hub with Robonomics Framework. The system of medical robots supports general sets of sensors and virtual devices that makes software-hardware interaction easy. Robonomics allows for decentralized cloud systems for robotics control on an open source Web3 platform. The existing public blockchain infrastructure and connects with Arduino's ROS ecosystem. This enables the exchange of medical information between humans and robots.
Medbot Robonomic's protocol, a peer to peer communication network is created for medical robot application. The peer to peer protocol and the use of smart contracts with uniform authorization protocols is transport agnostic allowing access to healthcare information for remote seniors. IPFS is a content-based distribution system and uses cryptographically verifiable hashes so this can store images like medical radiographs for medical diagonsis. This works by sending bio-metric data to IPFS and hash storing in chain options with Robonomics nodes. By using smart contracts and Blockchain technology, it provides secure integration of cyber-physical systems into the healthcare economy with seamless connectivity enabling distributed cloud technology for medical robotics.
Token-based authentication will help with security, where the token is a random IPFS hash assigned to the user and they can reset it at any point if it has been stolen. Allow the token to be passed in through the Medbot Robonomics network. Each transaction must be signed by account's unique seed. It has two forms: Mnemonic that is human-readable and Raw that is a sequence of digits and letters.
Enable medical robots to launch based on payment. Robonomics can be seen as a communication layer in ROS that dispatches the robots when the customer pays. It merges the technical information in ROS that makes the medical robot move and financing information into a single instrument. While all the infrastructure that enables this to happen is fully distributed, cannot be censored or controlled by any single entity. This builds trust among the healthcare services, provide direct medical access via dapp connected to decentrailized sensor networks.
Challenges we ran into
- Setting up IPFS and Robonomics
- Integration of custom hardware and cheap sensors into your ROS project using an Arduino. Using tutorials for setting up Arduino environment, creating a few sketches and hardware integration
Accomplishments that we're proud of
- Fall detection binary image classification algorithm was fun to make.
- Incorporating all the functions, methods, endpoints, requests, and responses from the Robonomics documentation.
- Solving remote healthcare addressing a large medical issue in society which affects many people daily, yet is often overlooked. Medbot robonomics solution can potentially contribute to the improvement of the medical system.
What we learned
- Blockchain + IOT + Arduino
- Learning about blockchain encryption with privacy sensitive bio-metric data.
- Issues in healthcare and aging in place
What's next for Medbot
- Go to market and how the decentralized application be available to the public, transformed into an MVP ( Minimum Viable Product)
- Add to Robonomics Framework with Robonomics IO - stream-oriented library with support general set of sensors and actuators (including virtual devices like PubSub or stdin/stdout) that makes hardware interaction easy
- Improve integration of IoT devices with legacy systems
- Research more about healthcare privacy and legal regulations
- Doctors will get notified to respective patient via SMS or Mail
- Multilingual Support
References
- 120 Artificial intelligence in palliative care: a systematic review to identify how novel data analytic technologies are used in the management of people with serious illness https://spcare.bmj.com/content/11/Suppl_1/A52.2?utm_content=consumer&utm_medium=cpc&utm_source=trendmd&utm_campaign=bmjspc&utm_term=usage-042019
- Artificial Intelligence for Caregivers of Persons With Alzheimer’s Disease and Related Dementias: Systematic Literature Review https://medinform.jmir.org/2020/8/e18189?utm_source=TrendMD&utm_medium=cpc&utm_campaign=JMIR_TrendMD_0
- Emerging Technologies With Potential Care and Support Applications for Older People: Review of Gray Literature https://aging.jmir.org/2020/2/e17286?utm_source=TrendMD&utm_medium=cpc&utm_campaign=JMIR_TrendMD_0
- EMERGING TECHNOLOGIES TO SUPPORT AN AGING POPULATION https://trumpwhitehouse.archives.gov/wp-content/uploads/2019/03/Emerging-Tech-to-Support-Aging-2019.pdf https://www.liveinhomecare.com/emerging-technologies-to-help-aging-americans-maintain-their-independence/information/
For my grandma.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent