The Intersection of AI and healthcare has tremendous potential for student projects. There is a wide range of healthcare sectors where AI can be used to automate and improve processes. In this blog, we focus on some key areas where middle school and high school can build impactful projects. We discuss hot topics in healthcare, possible projects, and examples of past AIClub student projects in these areas.
What are some examples of AI in healthcare?
AI can be used in any discipline that has data where patterns can be extracted from the data to make informed decisions. Healthcare is perfect for AI because it has a tremendous amount of historical and new data being generated every second. AI has already permeated several sectors in healthcare. Some examples are: drug discovery, personalized medicine, automated segmentation and analysis of medical images etc.
What are some hot topics in healthcare?
The table below lists some of the hot topics for AI in healthcare. These areas have many opportunities for kids to build innovative science fair or other distinctive projects using AI.
Intelligent Mobile Patient Tracking for Infectious Diseases
Sensors that track vitals such as heart rate, skin temperature and other physiological markers etc in addition to motion and activity are available easily in devices like smartphones, smart watches etc. Data generated from these devices over time can be used to detect the advent of diseases, impending heart attack and even COVID-19 days before any symptoms appear. There are several publicly available datasets that capture this type of information.
Foundational AI technologies for Faster Adoption of Telemedicine
Telemedicine is the practice of remote interaction between a patient and the professional health care provider for delivering care. Technologies such as video conferencing, messaging and AI are used to make this process smooth. The COVID-19 pandemic has a huge impact on health care as patients are afraid to visit hospitals due to the fear of exposing themselves to the virus. Healthcare professionals are also at a high risk when seeing patients at their clinics. Telemedicine is a practice that becomes necessary especially in such situations. AI can help the process by tracking, analyzing and alerting based on information provided from wearable sensors. For example, AI embedded into sensors can help doctors examine patient vitals without having to be present with the patient physically.
Disease prediction using multi signal analysis
Multi-signal analysis is another area with great potential. Instead of looking at data generated from a single device, combining data from different modalities is known to improve performance. An example of such data would be a combination of (a) clinical data such as height, weight, age, heart rate etc (b) MRI images (c) Histopathology images when a biopsy is ordered. Combining the rich and diverse information available about a patient over time allows the AI to detect complex patterns and interactions in data leading to better predictive performance.
AI Driven Remote Home Based Patient Monitoring
Remote home based monitoring of the patient becomes necessary in situations where the patient is unable to visit the doctors office, lives alone or has communication issues. During the COVID-19 pandemic, this has become much more common and necessary. This involves collection of patient data from sensors for monitoring heart rate, oxygen levels, blood sugar levels and then transmitting this to the doctors office regularly. Remote patient monitoring is going to become the norm and AI has a pivotal role in making this a seamless experience for both, the patient and the doctor. The amount of data generated by these sensors on a daily basis makes manual inspection by a professional healthcare provider prohibitively time consuming. AI can process this information and alert the healthcare provider with pieces of data that are concerning and need attention. This is an example of one of the many ways in which AI will become an integral part of this area.
AI Driven Medical Imaging
There are many types of imaging technologies in healthcare that enable doctors to diagnose and treat diseases, such as MRI, CT Scan, X-Ray, Histopathology images etc. However, there is a shortage in the number of experts who can read these specialized images. Experts can also take assistance from AI to speed up the analysis. AI has shown great promise in diagnostic medical imaging with impressive predictive performance. This is a key area with tremendous scope for innovation and impact.
Community Health
One of the key challenges in the current time is the risk and ability to go to a doctor. A very important question here is what can be done to improve the whole area of tele-medicine?. Think about the flow of what happens when you go to the doctor and how this process can be automated. Pick specific parts where the experience can be enhanced. This is an open ended question with a lot of possibilities.
Example healthcare projects by middle school and high school students
Here are some projects in healthcare built by students in middle and high school.
Forecasting Alzheimer's with AI and Multi-Modal Data
Danika Gupta, a rising 8th grader in San Jose, CA built a model for predicting the advent of Alzheimers. This project combines 3 different modalities to forecast Alzeimers (a) ADRC (b) MRI images (c) freesurfer. The dataset was from the OASIS and consisted of 30 plus years of longitudinal data about each patient. This is an example of rich publicly available medical datasets that can be used for building impactful projects and an example of the multi-signal problem outlined in the hot topics above. This project won a special award from the US Navy and Marine Corp at the Synopsys 2021 Science Fair. You can find more about this project here.
Co2 Sensei
Three students (Team MDS) built an iPhone app that businesses and individuals can use to measure the CO2 levels around them. Measuring this helps understand the ventilation level in any space, which is key to slowing down the spread of diseases. This is especially very relevant for slowing down the spread of COVID-19. They used a combination of hardware, AI and software for building iOS apps to accomplish this. This project has already been named a semi-finalist at the Technovation Girls 2021 Challenge. You can learn more about the project here.
Diabetic Eye Disease Diagnosing AI
Anika Pallapothu, a high schooler in CA built an AI to detect a diabetic eye from images of retinal fundus. Example images were obtained from a publicly available dataset. This will assist local doctors treat more people and can greatly impact the global treatment of the disease. This is a great example of both AI powered automation and AI-driven medical imaging. This project was named one of the 30 finalists across the United States in the prestigious Broadcom Masters Competition. You can learn more about this project here.
Breast Cancer Detection using AI
Shriya a high schooler from CA built an AI that can detect the presence of cancer from images of histopathology. Typically, a biopsy of the tumor is analyzed by pathologists to detect the presence of cancerous cells. An AI can speed up this process and assist pathologists to make speedy diagnosis. This project won the top award at Sandpiper middle school (the student was in middle school when the project was built). You can find more information about this project here.
Want to learn AI and build similar projects? Explore our courses!
Comentarios