Scanoma

Use your phone for checking your skin issues

13 followers

Use your phone for checking your skin issues

13 followers

Use your phone for a free skin health screening and get a second opinion from a dermatologist. Scanoma is a medical app combining computer vision technology and a platform for trained dermatologists working together to help reduce skin cancer risks.

Launch tags:iOS•Medical

Launch Team

EneyProactive AI for Mac. Free test, 20,000 credits.

Promoted

Great idea! I know it's hard for many people to justify going to the doctor for small things that have a tiny chance of leading to serious problems. Your app could fix that since it makes getting an AI's and a doctor's opinion painless. I can only wonder about the accuracy at this point though. What's your data source for the pictures used to train the algorithm and how accurate is it now? Also, when is it coming to Android? One more thing: the light gray color you chose for your body text (#B0B0B0) is REALLY hard to see against the white background (my eyesight is normal). Consider changing it to #757676 or darker (then it will also adhere to accessibility standards).

Report

7yr ago

Scanoma

Maker

@anna_0x Hi Anna, great feedback! The landing page is now updated with a darker grey for better readability and accessibility and we will do a full accessibility review to make sure the website is accessible and easily usable by everyone ! Regarding your other points: I 100% agree that for small issues medtech will be quite useful, at least in reassuring certain patients especially for issues that can be measured with constantly improving mobile sensors. Doctors have been telling me that sometimes patients even slow down the effectiveness of urgent care as some people go to emergency for minor ailments. Technology will not replace the skills and experience of doctors but can certainly complement it in some cases. Regarding the accuracy, it's much more than 70% in our set of test images, however, as with any AI classification tool, it can perform much better or much worse depending on the images and conditions it runs on (lighting, obstructions, similarity to the images the algorithm was trained on...). Our training data is a combination of high quality academic-grade research images augmented with our custom constructed images to take into account things such as different skin shades and conditions. I hope this answers your questions and thank you for your support ! Android version is coming soon. I will be happy to send you a notification when it's released.

Report

7yr ago

@nanomeko it’s readable now! I assume that the processing happens on the device itself since it works without a network connection, right? Maybe mention that somewhere on the website? Speaking of the website, consider using SVG graphics (which are infinitely scalable) instead of plain old image files, or at least use high definition images. They look really blurry on Retina or similar displays—especially the logo. Also, the privacy policy and T&C strangely seem to refer to using the website and not the app (which doesn’t even include a link to either). I say strangely because your product is an app. The website’s only purpose is to promote said app. The privacy policy covers the website, which doesn’t really do anything, but not the app, which lets people upload potentially sensitive photos. A very weird decision to say the least.

Report

7yr ago

Scanoma

Maker

@anna_0x Thank you for the feedback! We will definitely replace the few non SVG images with scalable versions. Regarding the privacy policy & TC's it's one of those things which I put on the low priority column until we get a professional lawyer to write better ones soon. We take users' privacy extremely seriously. All images are sent anonymously to doctor and a lot of effort is made to make sure everything is set-up correctly in regards to anonymity and security. However, the website does more than it seems. The product is actually the scanoma platform combining the mobile app and the website. The website is the platform for doctors to review patient images and give their professional opinion. The mobile app is just a starting point for the end user with the goal of getting a professional opinion from real human doctors. We will modify the privacy policy and T&Cs to reflect that. Thank you again for the great feedback !

Report

7yr ago

I hate going to the doctor and I hate making appointments— this could be awesome if I can use it without having to have health insurance.

Report

7yr ago

Scanoma

Maker

Hi, I'm Nouaman maker of scanoma. I hope this platform makes it easier for everyone to have easy and affordable access to medical professionals for skin health issues.

Report

7yr ago

This is a wildly inappropriate app. I just scanned 3 moles and got either 86% benign or 70+% concerning depending on the angle, flash etc.

Report

7yr ago

Scanoma

Maker

@psychometric Hi Jason, thank you for your feedback. As with any ML based classification tool, you will receive better results if you scan the moles in the same conditions the training images were taken in. We encourage users to use the results as an indication of concern and always send their images to trained dermatologists for a second opinion. We believe and encourage users to understand that technology complements human medical skills but does not replace them. Thank you for your feedback!

Report

7yr ago

No-Code Tees by Maker Threads

Such a valuable product. Would love the know what level accuracy the model is able to predict.

Report

7yr ago

@lachlankirkwood depends on your training data i believe. I mean you could have inaccuracies in your training data that wouldn't be accounted for in your model.

Report

7yr ago

Scanoma

Maker

@lachlankirkwood I’m happy you see the value in the product. We think it will be especially useful in remote areas with less access to medical care as a sort of first layer of care. We always recommend users to get their skin issues physically checked by a human though. While accuracy is pretty high, lots of things can mess with it (for instance tattoos). This can be solved by the continuous training with more data. @robbyreindeer Exactly. Good training data would result in good accuracy. The training data needs to also be similar to the real life situations (lighting, camera types, etc.)

Report

7yr ago

Form-Builder

Sounds like an extremely valuable app. Waiting for the Android version to check it out

Report

7yr ago

Scanoma

Maker

@danbars Glad to hear you see the value in the app ! Android app is in the planning stages. We wanted to see user reception before building a native Android app. I'd be happy to send you a notification when it's released !

Report

7yr ago

Reviews