NIRSTest1

The project mainly consists of 2 parts: Android part and Machine Learning.

This is the research project by me and Chu Luo, there are just some previous versions of our work for demo, please contact us if you have any interest!

========================================================

1. Objective

Smart medication: See there are 2 scenarios we can make use of the technology. One is for elderly home medication, as old people take lots of pills at home, they may be unable to distinguish the pills clearly with boxes. The smart medication can help them scan the pill and give proper instructions. For example, when old people live alone, they do not have literacy or see the words on the pill box clearly, also their children should work and have no time to take care of them during the day. The children could set the app for them in advance, and the app could have alarm function, it would remind the elder to take pill A in the morning, pill B in the noon (after lunch), and pill C after dinner. Besides, it can give clear and big signs shown on screen, indicating the right or wrong pill they are trying to take. The other one is for hospital emergency. About 125,000 people die every year due to medication mismanagement and the estimated cost is around $300 billion. Miniaturised NIRS could be implemented to help nurses or the users themselves (e.g., old adults) administer medicine [1]. E.g. when the new nurses are not so familiar with the medicines, not good at English, or they are busy or tired at night. This technology can be a good helper for picking the proper medicines, and reduce the probability for mistakes.

2. Background

Near Infrared Spectroscopy (NIRS) is a technique that propagates near-infrared waves through objects and measures the magnitude of the reflected light across the spectrum. This technology enables users to infer highly detailed information about objects. Traditionally NIRS has been an instrument reserved for laboratory usage, which is large and expensive. But recent developments have caused the price, size and weight of these devices to decrease. Pairing the Miniaturised Near Infrared Spectroscopy with smartphones can open up a plethora of new use cases.

A NIRS device used in our study (DLP NIRscan Nano) costs under 1000 US dollars at the time of writing and weighs just 80 grams -- a fraction of the price and weight of high-end NIRS hardware. This dramatic drop in cost enables us to consider everyday scenarios for this technology, and for the first time put it in the hands of non-expert end-users. Coupling NIRS hardware with commodity devices (tablets, smartphones) opens up a range of exciting research avenues to explore. An in-situ scanner placed at an elderly care centre would allow for classification of pharmaceuticals before they are distributed to the users. In such a scenario, a user can scan a pill or medicine to confirm that it is the right one to take at that moment [2].

NIRS have ability to scan the infer compositions in objects, because different objects have different compositions, we will get lots of spectrums after the experiments, and get big data for distinguish the objects with the waves.

Thanks for Simon Klakegg, Jorge Goncalves, Niels van Berkel, Chu Luo, Simo Hosio and Vassilis Kostakos, who finished a comprehensive technical research on this area [1]. At this point, I plan to extend the topic based on their work.

REFERENCES

1. S. Klakegg, J. Goncalves, N. van Berkel, C. Luo, S. Hosio, V. Kostakos. 2017. "Towards Commoditised Near Infrared Spectroscopy", Proc. Designing Interactive Systems (DIS), pp. 515-527.

2. Candolfi, A., et al. (1999). "Identification of pharmaceutical excipients using NIR spectroscopy and SIMCA." J Pharm Biomed Anal 19(6): 923-935.