Using Artificial Intelligence to distinguish between suicidal and depressive messages
We are an initial screening service that efficiently helps concerned friends and family recognize depressed and suicidal tendencies in their loved ones. Using Artificial Intelligence and Natural Language Processing, we build on existing products to come up with a unique and unprecedented solution to this pressing issue.
Please note that this application is still in development, and no prediction should be taken seriously. Refer to clinically-verified professionals for mental health diagnoses.
Those with mental struggles tend to first open up to trusted friends and family. These people can express themselves to loved ones rather than online robots or public social media, which makes them more successful in combating depressive and suicidal tendencies. As communication is increasingly online, SuiSense strives to help you understand your loved one’s mental state and recognize possible flags through online phrases and texts.
While depression and suicide are interconnected, recognizing and treating each have significant differences. SuiSense is able to efficiently differentiate depressive ideations from suicidal tendencies, which can lead individuals to get psychologists who understand how to most effectively approach treatment. In addition, our product helps family and friends understand more clearly what their loved one is going through.
COVID-19 has created countless additional problems in our lives, and many of the pressing concerns have been addressed in some way. However, not many solutions have addressed the impact of COVID on mental health. Depressed and suicidal individuals must deal with health problems, employment issues, and social isolation on top of their already fractured mental health. Depression and suicide have increasingly marginalized those struggling with mental health; in the Bay Area, we have seen more deaths from suicides than COVID-19 during quarantine. We aim to alleviate some of these struggles through SuiSense.
To have a strong community presence, we believe partnering with organizations and therapists can make our solution as precise and effective as possible. Through the help of certified therapists from the Bay Area, we have been able to tailor our solution to help those suffering from mental health issues in the most effective and unique ways possible. Through partnerships with organizations, we have been able to promote our solution and begin the implementation process. We aim to create a community where individuals are not afraid to reach out and gain support for their mental struggles.
Using our initial NLP model, friends, family, and therapists can paste or write out text, and will be able to determine if their friend is suffering depression or is on the suicidal side. This provides important initial screening which determines the fine line between depression and suicide.
To personalize our screening to each patient, we intake previous messages from before a patient indicated a change in mood or attitude. This allows our models to make a personalized diagnosis based on their previous state and determine how much they have changed over a period of time.
Our progression model which classifies the patient into one of 6 different categories based on Jessie Bering's classification which gives a more specific diagnosis for the patient. By intaking multiple messages, we are able to determine which stage a patient is on with a psychologist-certified scale.
Friends and family are able to visit the local map we provide to find therapists for their loved one. Using our pairing algorithm, we are able to pair individual patients to therapists that fit their needs the most, based on parameters provided by our certified therapist consultants and advisors.
We use Natural Language Processing and AI to tackle problems with diagnosing depression and suicidality. Our solution provides key insights into patients based on the phrases they say, which has been proven to be an effective way to diagnose depression and suicidality, but with NLP, we leverage data to determine the fine lines between the two.
Throughout development, SuiSense has consulted with these psychologists with professional experience to understand the issues facing depressed and suicidal teens.
Dr. Ruebsamen is a licensed Bay Area psychologist who founded the "Always Be Listening Global Initiative", a global humanitarian psychological project working with adolescents and families. She is also part of Abbey Blue Canoe Open Path Collective, a United Nations organization for uprooted children. As our advisor, Dr. Ruebsamen helped us identify our target group, refine our process to be as effective as possible, and improve outreach and implementation. In addition, she is connecting us with potential users and assisting us in implementing our algorithm.
This is an informational video demo of SuiSense submitted to the Congressional App Challenge (awarded 4th place). Watch to see SuiSense in action.
As an extension of our initial project, with the novelty of our method in mind, we wrote a paper on using deep learning for suicide vs depression identification, and the model is named SDNCL. Importantly, we address the problem of noisy labels using unsupervised clustering. SuiSense now implements and uses the models from SDCNL.
SuiSense proudly has built partnerships with local student-led organizations in order to maximize our impact and implement our solution.
The Children's Addiction Prevention Association (CAP Association) is a 501(c)(3) nonprofit that hopes to prevent drug abuse among teenagers. Their process hopes to provide the future generation with the tools for careful decision-making. Dr. Ruebsamen indicated that drug abuse is strongly related with depression, so we sought out to build a partnership with CAPA, as our missions align. CAPA has promoted and supported SuiSense, and the user base of SuiSense has grown as a result. CAPA has presented SuiSense at their events, and neighboring schools have begun showing interest.
The Mental Health Initiative (MHI) is a Bay Area organization dedicated to eliminating the single story surrounding a typical bay area student stereotype. MHI is highly motivated in dismantling the stigma around mental health problems in the Bay Area and bridging the gap between parents and students regarding important mental health problems. To promote mental wellbeing, we established a partnership with MHI in order to increase our impact and user base. SuiSense has been supported and promoted at MHI’s local events and meetings, such as those at the Saratoga Library.