Most doctors make diagnosis based on their personal experience and the research and publications they expose themselves to. Could they have had different diagnosis if they were more aware of data they did not get exposed to? What if machine learning could be used to offer diagnosis based on all current and constantly updated data (from patient cases, research findings, publications and other doctors' experiences)?
Also, most doctors can't trace the exact reasoning behind the diagnosis. What if they could?
Kahun came with the idea to do just that and needed a way to materialize it into a product and a brand.
Build growth strategy on multiple fronts and prioritize what's first
Build a brand
Create an efficient technical tool that shows the reasoning behind symptoms and possible diagnosis based on constantly updated medical ai (a beta app)
Research, strategy, brand, product
Build growth strategy on multiple fronts and prioritize what's first
Build a brand
Create an efficient technical tool that shows the reasoning behind symptoms and possible diagnosis based on constantly updated medical ai (a beta app)
Research, strategy, brand, product
Group A - medical students (early adopters)
Group B - doctors (professionals)
Logical, technical, transparent, fast
Group A - medical students (early adopters)
Group B - doctors (professionals)
Logical, technical, transparent, fast
Group A - medical students (early adopters)
Group B - doctors (professionals)
Logical, technical, transparent, fast
Group A - medical students (early adopters)
Group B - doctors (professionals)
Logical, technical, transparent, fast
Medical diagnosis thinking in its core is all about statistic routes.
The needs of a medical student is to get out his phone in a moment of a blackout and get a clear medical reasoning route between simptoms and diagnosis under 1 minute.
The need of pro doctors is to get the bottom line knowledge of the cases they are not aware of that can effect their diagnosis.
Overview of a combined routes map can provide additional insight to the doctors but does not suit a mobile device for its spread, so it won't be included in the beta.
Translating medical texts into statistics equations used by the ai takes a lot of manpower with medical background.
Use community's versatility - encourage users to be data contributors of the app by vetting them and crediting them for new entries of updated medical case studies and articles.
Connect users to latest research - notify doctors of freshly updated sources of knowledge of the field thay practice/want to follow.
Use it as a teaching tool - give option to manually add diagnosis that was not initially recommended and the student might want to check, just to show its reasoning route has low probabilities.
Show credibility by sharing the bibliography sources for each reasoning map.
Use community's versatility - encourage users to be data contributors of the app by vetting them and crediting them for new entries of updated medical case studies and articles.
Connect users to latest research - notify doctors of freshly updated sources of knowledge of the field thay practice/want to follow
Use it as a teaching tool - give option to manually add diagnosis that was not initially recommended and the student might want to check, just to show its reasoning route has low probabilities
Show credibility by sharing the bibliography sources for each reasoning map
Kahun is the name of an egiptian medical papyrus, one of the largest ever found. It being a source of all known medical knowledge back then was the inspiration to nowdays Kahun's mission - to be a united source of the most updated and accesable medical knowledge of today. The logo is a book, an illuminator and an arrow to point to a clear answer to anyone's searching.
Initially the app was divided to 2 sections: solve and view. Solve is for solving a medical case in 3 steps: 1.findings 2.diagnosis 3.workup (recomended course of action = medical tests, which results are necessary to weaken or strengthen certain diagnosis). View is for getting recent data updates from Kahun's library of medical resources by date and field.
Testing it showed that for the beta, the focus should be on the "solve" section, discarding the "view" section. The medical bibliography would stay as an inner link only. Another discarded step was "workup", to make the beta more inforamational than a step by step guide.
The remaining stages of findings and diagnosis - since being equaly important, as they constantly effect each other - were advised to be visible simultaneously.
Medical reasoning, that prior showed which diagnosis is "weakened" or "strengthened" by which finding, was asked to be shown in a more detailed way, making a higher emphasis on it being the main gem of the app.
Entering finding is the departure point of the user journey. The arrival point is the immediately updated diagnosis probabilty section. Pressing on the diagnosis shows the clinical reasoning that connects the two points.
The main screens of findings and diagnosis are dark, inspired by the common dark interface of technical medical equipment, while the details of the medical reasoning and its sources are bright, inspired by the A4 print out data doctors are used to holding in front of them.
A product that promotes wellness and healing was added to the world, the beta was successful, helpfull feedback about the app was collected (for ex. adding more data input gradualness and approachable complexity by adding more sections and icons), new changes were made and now it's an active google app store app.
Research and analysis, growth strategy, logo + brand language + brand materials design, motion design, clinical reasoning mapping language solution, beta app ux ui
Michal Tzuchman Katz (medical consult), Eitan Ron (feedback and funding)
Research and analysis, growth strategy, logo + brand language + brand materials design, motion design, clinical reasoning mapping language solution, beta app ux ui
Michal Tzuchman Katz (medical consult), Eitan Ron (feedback and funding)
Research, strategy, logo + brand design, beta app ux ui
Michal Tzuchman Katz (medical consult), Eitan Ron (feedback and funding)
Selected Works
Site SelectionProject type
KahunProduct, Brand
WavesStrategy, Product
YoProduct
EngieProject type
LeomCampaign
Qiryat YovelProject type
RachminovWeb, Brand
Euro EstatesProject type
Russell'sWeb, Brand
BliinderProduct
Humanities FundWeb, Brand
Urban EncountersProject type