Melanie, Your Period Fairy Godmother in an App
Note: Melanie is not intended to replace a doctor by any means. We strongly encourage individuals to report any symptoms to a medical professional.
Periods.
A very taboo topic.
But why?
Periods and cramps just aren’t talked about enough in society. Contrary to popular belief, menstrual pain is 100% real and affects the majority of women all over the world. To this day women around the world still face severe inequalities in regards to finding access to adequate resources, advice, and help.
According to a national study conducted among over 40,000 women aged 15–45 in the Netherlands in 2019, 85 percent of women reported having dysmenorrhea, or period pain, and 38% of the women reported not being able to perform their daily activities.
In addition, from the women that had to skip tasks or activities because of their symptoms, only 48.6% of women told their family that menstrual symptoms were the reason for their inability to complete what they were doing.
Let’s think about that for a second.
Not only are more than 1 in 3 women experiencing discomfort to the point where they can’t carry out routine activities from going to school to taking care of their children, but more than half of them are unable, to tell the truth to their family and friends.
This is just in the Netherlands.
In India, menstruating women aren’t allowed to cook, pray, or eat certain foods, and are considered “dirty” and “impure” during this time.
In China, approximately 2% of women use a tampon as women are told that using a tampon takes away one’s virginity.
In Kenya, one in four girls does not know that they can get pregnant once starting their period.
Even in the United States, a study commissioned in 2018 found that 58% of women feel embarrassed simply for being on their period, 42% of women have experienced period shaming, and 12% of women have been shamed by a family member.
Periods aren’t seen as inferior in just one country. This is a global issue. And women deserve better.
Not only is there a horrible stigma around periods, the stigma funnels into more serious issues for individuals. There is an incredible lack of resources and guidance for people not just in the U.S., but around the world. Women struggle to find the right people to talk to and understand which resource is giving them the right advice. Many women are told to talk to their physician about issues with their menstrual cycles. So what is the issue with this?
For starters, how often do you really see your doctor? Every six months? A year? Going to a doctor isn’t a routine practice in developed countries, much less in undeveloped ones. Long breaks between visits means that patients cannot explain small occurrences that could have a long impact on their reproductive health over a longer period of time.
Let’s lay out a very typical scenario that a large portion of adult women face. You go to see your physician at your yearly checkup. You bring up the issue of your agonizing periods. You can’t really remember how long your typical period is, what the flow looks like, or much else, but with vague detail, you describe your bloating, terrible cramps, and several occurrences of dizziness or even fainting. Your doctor deems that you have a “bad period,” or have “imbalanced hormones,” requiring you to go on birth control immediately while never explicitly addressing the complications that come with it.
In an online survey conducted by the National Pain Report and For Grace, a nonprofit organization, researchers found devastating results. In a survey conducted among 2,400 women, over 90% of individuals felt that the healthcare system discriminates among female patients. 84% feel they have been treated differently by doctors because of their sex, and 65% feel doctors take their pain less seriously because they are female.
The lack of information, misguidance, and stigma even at the doctor’s office can result in a buildup of consequences along a woman’s lifetime.
Nowadays, more resources have sprung up on the internet to “diagnose” individuals by listing trademark symptoms, influencing women to take certain drugs to help with cramping or other menstrual problems or give their opinion on the optimal lifestyle for women’s health. Ads and sponsored articles for products that claim to reduce period cramps and acne or blogs advising certain practices show up wherever we look.
In addition, “femtech” has become a buzzword mostly in developed countries. We’ve seen the rise of apps such as Clue and Flo, which promise to track menstrual cycles, ovulation cycles, and mood through manual user input and give access to other resources about birth control, menstruation, and fertility, to name a few topics.
However, these apps aren’t all they seem to be. They involve “premium memberships” and include heavy promotion in an attempt to generate revenue. Paywalls, ads, and limited features keep many women from having access to the entirety of the resources they deserve.
We’ve addressed the problems and stigma about periods around the world and the issues with currently available resources. So what is the solution?
Imagine if all the answers to all of the questions women had were in one app. The answers to uncomfortable questions. Questions about periods, activity, lifestyle, or even biology. The answers to awkward and confusing questions. An app where women could not only track every aspect of their menstrual cycle but could openly discuss their concerns and ask for personalized and honest advice.
This is exactly what we are proposing.
We are proposing Melanie, a no-cost virtual assistant that communicates through speech to provide answers to questions about reproductive health and lifestyle, from “is it normal to skip a day on my cycle?” to “why is my blood darker than usual?” to “how much water should I be drinking to minimize bloating during my cycle?” Melanie can answer virtually any question the user has about period irregularities and give advice for the user to optimize their lifestyle, reducing pain and symptoms throughout the entirety of the menstrual cycle.
So what exactly would this app look like?
Melanie is an app personalized to fit the user’s needs. The user has the ability to utilize Melanie as little or as much as she wants. The user has the option to manually complete a simple daily check-in logging their physical and emotional well-being that day, or they can input symptoms whenever they feel necessary. The user can track their menstrual cycle, as well as…..
- cramping
- appetite changes
- fatigue
- changes in skin
- spotting
- bloating
- breast pain
- constipation/diarrhea
- headaches
- dizziness
- lower back/pelvic pain
- mood changes
- nausea
The user can track these symptoms not only during the menstrual cycle but on any given day. The user can also input dietary changes, water intake, or changes to their sleep schedule. The user has the choice to manually input these aspects of their cycle or tell Melanie to track them. A simple statement after tapping Melanie, such as “I am bloated today,” will easily input the data for you.
Our app will contain two main interfaces. On the homepage, the user will see a calendar, similar to a normal period tracking app. It will have the days of the week as well as a countdown towards the projected start day of the upcoming menstrual cycle. When the user clicks on a day of the week, they will be taken to a page where they can log the information listed above. Melanie will be able to access this information in order to give recommendations. On the bottom of the screen, there will be a bar with the option to switch to the voice assistant interface. When the user clicks on that, they will be taken to a screen that has the option to either do a daily check-in or talk to Melanie.
In the daily check-in, the user can respond to questions such as “Are you experiencing headaches, nausea, pain, etc.” or “Are you experiencing any external stresses? (alcohol, travel, stress, change in food intake/nutrition, etc.)” by selecting options out of a multiple-choice list. This allows the user to easily input large amounts of data in a concentrated amount of time, allowing for more accurate recommendations and advice.
The user has the option to ask Melanie about any concerns or for advice at any given time. There will also be an option to type a question if the user is in a situation where they cannot speak out loud. If at any time Melanie notices an irregularity in the data the user presents, she will give the user a warning screen that will display the irregularity and a way to learn more. These irregularities could be anything from a late period to a sudden drop in sleep or water consumption.
Not only will Melanie track all of the user’s data in one place, she can make predictions not just about the user’s menstrual cycle, but also about symptoms they might experience. If the user tells Melanie they are dealing with migraines 2 days before their period, Melanie will be able to remember and record this information. As this pattern begins to happen more consistently, Melanie can begin to warn the reader two days before their predicted cycle that they may be experiencing this certain symptom.
Unlike the resources described earlier, Melanie is also transparent and personalized. What does this mean?
Melanie gives the user honest and open advice. She eliminates any of the stigma found at the doctor’s office and misinformation is written on the internet. Melanie gives you a straightforward answer and all of the options and recommendations in a certain scenario.
Let’s say the user asks Melanie why they have been so bloated on the first two days of their cycle. Melanie can go back to the user’s history and note that they haven’t dealt with bloating over their past 12 menstrual cycles. However, she might notice that the user’s average water intake over the past week has been slightly lower than it usually is at that phase in the user’s cycle. She may also notice that the user has been more sedentary than they usually are at this phase in their cycle.
By finding minor differences and details throughout the user’s current and previous cycles, Melanie would present the user with a list of recommendations. However, Melanie acknowledges the fact that it could be a more serious issue. She will always recommend that the user asks for further guidance from a medical professional, whether it be a physician or OB/GYN. She will also ask the user to continue to take note of the bloating to see if the issue persists, assuring that it is addressed so it doesn’t become a long-term issue.
Melanie will always address all sides of the problem and will always present the user with all of the options. She gives her advice and recommendations for small changes the user can make, asks them to take note of certain areas to see if problems persist, and always encourages the user to seek further help to prevent skipping over something that could persist and develop into a much larger issue.
Not only is Melanie transparent with the user, she provides the user with a completely personalized interface. Currently, no app or resource has been designed or created to give each individual personalized advice in regards to their menstrual cycle. Doctors cannot collect enough information about a patient over a period of time to give advice on an issue or observation from the patient. Blogs, websites, and apps are created with the “general population” in mind.
But every single woman has a different experience.
No one woman lives the same life as another. We are all exposed to different internal and external factors, whether it be our genetic composition or our diet. One woman might experience horrific cramps while her twin sister experiences nothing. One young girl might faint on the first day of her menstrual cycle every single month while her mother has never had any sign of dizziness on her cycle. We cannot generalize anymore; we must start addressing the diversity of experiences women face.
Melanie is able to create such personalized recommendations from the input she receives. If patients are asked to manually input their data themself, they are more unlikely to do it. By creating a simpler interface that allows the user to just speak, users find that they have less of an excuse to log symptoms or changes. In addition, most algorithms likely won’t pick up on the root cause of irregularities. Apps such as Clue and Flo are trackers. They aren’t designed to notice patterns but expect users to look back on logged data themselves. However, Melanie acts as a resource to cut out this step. She has the ability to pick up on subtle patterns and find opportunities where the user can modify their lifestyle to reduce menstrual symptoms and disorders.
So how exactly does Melanie work?
Our app contains two aspects of AI; natural language processing and neural networks. Melanie uses natural language processing to understand what the user is saying and report back.
Let’s look at how these technologies work within Melanie by asking her a question: “Hey Melanie, my cramps are really bad, what should I do about it?”
At this point, our question will travel through a series of stages to be broken down and understood by the software. Then, the topic will be searched up in our internal database to find answers and be reported back to the user.
- Tokenization — This occurs when the software takes a big chunk of speech and splits it up into smaller chunks of words and sentences. These smaller pieces are known as tokens. In this situation, Melanie would break the question down into tokens of “my,” “cramps,” “are,” “really,” “bad,” “what,” “should,” “I,” “do”, “about”, “it”.
- Stop Word Removal — Stop words are common articles, pronouns, and prepositions such as “I”, “really” and “my”. Most of the time, these words offer little to no meaning or value in the context of the sentence and it is easier for the software to get rid of them.
- Lemmatization — This is when the software cuts out parts of a word so it becomes a root word or “lemma” that is easier to recognize. For example, the word “running” would turn into its root word “run”.
- Named Entity Recognition — The goal of NER is to label noun and noun phrases with the concepts that they represent. The system uses context clues to create a “definition” for each individual word. For example, the computer can recognize the word “cramp” as the concept it represents.
- Coreference resolution — This is the final and most complicated step. Pronouns are mapped across sentences and words that refer to the same topic are connected. If a human reads the sentence “My cramps are really bad, what should I do about it?”, they know that the “it” refers to the cramps, however, the software would not know that.
- Information Retrieval — The software will access our own internal database to retrieve information regarding the specific questions. In this case, the software will look up information regarding period cramps, possible reasons for pain, and possible remedies. The software will come across multiple “answers” in the form of broader concepts for this issue ranging from dehydration to more serious medical conditions such as uterine fibroids.
- Finding “Trends” and Generating A Response —
To make Mel extremely accurate, we will collect vast amounts of data to develop our own internal semi-public database. We will collect information from the NHS and other researchers and OB/GYN professionals to input in our database.
With knowledge of these possible reasons for your cramps, the software will interpret data submitted by the user that is stored within its interface. It looks back at data received through manual and verbal input over the user’s history. The software will then analyze the user’s data to formulate a “trend” anywhere from the past few months to year, depending on the amount of time the user has regularly been inputting data. For example, if the software looks through the database and finds dehydration as a cause for menstrual cramps, it will find a “trend” for the user’s water intake every week over the past year (assuming that the user had used the app for at least one year). It can then compare the user’s water intake for this particular week to all past weeks over the last year. If the software finds a statistically significant difference (ex. A 10 oz difference in average water intake) it would formulate a recommendation for the user.
However, the software might not notice a difference or might notice a statistically insignificant difference (ex. 2–3 oz). Then, it would move on to investigating trends for another possible cause of cramps. For example, if it finds that uterine fibroids could be a cause of cramps on an database, it would investigate the symptoms of this medical issue and formulate trends for these symptoms. These symptoms would include extremely painful cramps and a very heavy flow. The software would then create trends for the user in regards to these symptoms over the past year and evaluate the probability that the user could be facing a more serious medical issue. If the software finds that there is any chance that the user has uterine fibroids, it will alert them immediately and strongly encourage them to seek the help of an OB/GYN or medical professional.
So what happens once Melanie receives the answer from the database?
The answer is transmitted to the user as a recommendation. If Melanie has found that the cause of an issue is minor, such as dehydration or lack of sleep, she will personalize a specific recommendation to the user. For example, if Melanie finds that the user needs to sleep for two more hours to reduce migraines in relation to their menstrual cycle, she will show them the recommendation on her recommendations page shown below.
What’s Next
Melanie has the potential to transform the lives of women around the world. Women in developed and developing countries, young girls or adults, women of all backgrounds, cultures, and stories. Melanie is designed to represent and serve all of these women.
Melanie has the potential for impact as it eliminates the bias and stigma found around periods and menstrual cycles in social settings, even at the doctor’s office. Melanie provides an interface where individuals feel that they are getting the help they need without any judgment. By creating this resource, we hope that individuals across the world will be able to find answers to their uncomfortable, confusing, and personal questions once and for all.
Thank You!
Thank you so much for giving this article a read. If you enjoyed it, we would love it if you could give it a clap! Also, please connect with our team below!
