Team Oto – Build progress!


The design and progress of Oto – the digitised otoscope, has been really good till now. The team has been busy with improving the wearablility and ergonomics, and concentrating on getting the imaging right.

Here are the parts that we are working on, along with the challenges:


We have decided to 3D print a cone that would cover up the digitised otoscope, provide control over the focus of the lens inside and plus may be incorporate a switch which would in-turn illuminate the inner ear. The print would be done by today evening, hopefully.

One of the challenge would be to get  weight balancing right for the head-phone-like wearable, may be by using a strap tp hold it in place or using the battery power source weight on the other ear to keep the contraption steady.

Another challenge would be to keep the speculum (the nozzle) steady and in place to get good images. The movement / positioning of the speculum should be defined well in-order to fit the same into the ranging shapes and sizes across many patients, from kids to adults. Hopefully the cushioning on the outer-ear would provide some freedom to move around and reach the corners, to capture quality images.


We have used a combination of the otoscopes original lens, combined with a telephoto lens to capture the ear drum and the canal. The stripped webcam captures this image and provides it to a laptop for further processing.

One of the challenge is that there is a lot of reflection from the inner ear that’s making the images foggy. Another one is that the focus getting lost in the process. But with better alignment, build quality and more control in the focus we will get closer to capturing good images.


Krishna testing out Oto.

Ear drum image 1

A sample image of the ear drum captured.

Image processing:

The images captured needs to be processed to find out variations in the ear drum and surrounding areas to detects Otitis Media and other disorders too, later. There was a discussion in how to go about it. The idea is, li,e usual, to detect the edges and contours on the ear drum and identify the prominent parts to match it to a standard template. And then with the absence / presence, and even partial presence of some of the parts, along with color variations in the ear drum, you would be able to tell if someone has an ear infection.

Some challenges are that we need lots and lots of images to learn and do processing to find out general variations in the contours and color of the ear drum among people before and after they are infected. Eventually the machine would learn from the increasing data set. For that one of the core components is to get the imaging part ready so that we can get more images to play with.

And we are glad that Vaibhav with his immense 3D printing knowledge and quick design sense has joined our team!

More updates soon!

Take care,


Team Oto: Anshuman, Vijay, Krishna, Neeta, Surbhi, Foram, Manu and Vaibhav.