Fitness Tracker Hacked Into Optical Density Meter: A DIY Project for Biohackers

The generic fitness trackers usually contain a microcontroller, BLE radio, photodiode, rechargeable battery circuit all stuffed in a tiny package. Therefore, the entire device can cost less than $25 and this fitness tracker based device, ODX, can be used as the cheapest continuous OD meter.

Fitness Tracker Hacked Into Optical Density Meter

DOWNLOAD

The lingering uncertainty in the quality and usability of the underlying sensor signal quality motivated our work here, by first evaluating the raw IMU sensor data quality of smartwatches, and then second, trial the feasibility of large-scale deployment in a clinical care setting through the patient (PPI) and healthcare worker involvement. Clinical and care wearable applications and analysis rely upon quantified, regulatory acceptable measures of accuracy of the fundamental signal (linear acceleration for accelerometers and angular velocity for gyroscopes) that must be compared to the reference standards. The quality in these fundamental signals allows us to assess how well they can in principle track measures, such as body kinematics, but also more indirectly inferred measures, often clinical outcome measures and primary endpoints of clinical trials (such as step counts). To date, there has been no independent direct comparison between common consumer smartwatches, research-grade IMUs inertial and ground truth optical motion tracking. This is in part due to consumer smartwatch closed-system barriers to raw IMU data extraction, which we overcome through developing customised software. We also developed an easily reproducible measurement protocol to directly assess and compare smartwatches in naturalistic movement tasks, performed by the same human on all compared devices at the same time. Additionally, a separate issue for the clinical feasibility of smartwatches in care is the feasibility of their deployment and their practicality and acceptability in everyday use by both patients and healthcare workers. Research to date on patient and healthcare staff attitudes towards the continuous wearing of IMU sensors is scarce. While some studies report user-perceptions (e.g., user-friendliness and satisfaction) of smartwatches and fitness devices [20,21,22,23,24,25], these often focus on community settings, chronic disease, young/middle-aged subjects, and healthy participants and, as such, are not as relevant for typical in-patient populations.

Thus, visual acuity, or resolving power (in daylight, central vision), is the property of cones.[21]To resolve detail, the eye's optical system has to project a focused image on the fovea, a region inside the macula having the highest density of cone photoreceptor cells (the only kind of photoreceptors existing in the fovea's very center of 300 μm diameter), thus having the highest resolution and best color vision. Acuity and color vision, despite being mediated by the same cells, are different physiologic functions that do not interrelate except by position. Acuity and color vision can be affected independently.

Widefield microscopes can be adapted for spectral imaging using a technique known as Fourier transform imaging spectroscopy by coupling an interferometer to a fluorescence microscope. In this configuration, light gathered by the objective is directed to the interferometer, which splits it into two independent beams and simultaneously introduces a minute optical path difference (OPD) with a slight time delay between the beams. After passing through the microscope optical train and specimen, the beams are subsequently recombined to create an interference pattern that is projected onto the digital camera (CCD) detector. The resulting intensity pattern of varying OPDs at each pixel is termed an interferogram, which is specific to the spectral content of the specimen. The image stack can then be treated with a fast Fourier transform (FFT) algorithm to determine the profiles and spatial locations of the contributing spectra. Among the advantages of this methodology are that no filters are required to separate emission because the spectrum is measured using the interference of light, and the intensity at each wavelength is collected for each image captured during the process. Additionally, the spectral resolution is determined by acquisition parameters and can be changed without modifications to the hardware configuration. On the downside, the full spectrum must be collected for each image even when only a limited region is required for the experiment. Fourier transform spectroscopy is widely employed for spectral karyotyping and has also been applied with linear unmixing to separate emission from colocalized fluorescent probes.

A typical spectral image lambda stack gathered by the microscope is often composed many thousands or even millions of individual spectra (depending upon the lateral image dimensions), with essentially one spectrum being represented at each pixel location. The associated data files are therefore quite large and complex (virtually impossible to analyze by visual inspection), thus requiring a dedicated software palette for interpretation and presentation of the results. Analysis of lambda stacks can target the extraction of spectral data or image features (or both) using tools that are either packaged with the instrument or widely available from aftermarket manufacturers. Each fluorophore or absorbing dye, regardless of the degree of spectral overlap with other probes, has a unique spectral signature or emission fingerprint that can be determined independently and used to assign the proper contribution from that probe to individual pixels in a lambda stack. The result of the linear unmixing technique is the generation of distinct emission fingerprints for each fluorophore used in the specimen (or excitation fingerprints if excitation rather than emission spectra were employed to generate the lambda stack). In summary, the spectral information in an image captured by the microscope is binned into three broad spectral ranges roughly corresponding to the primary additive colors red, green, and blue. Linear unmixing enables the precise determination of spectral profiles at every pixel in the image to overcome overlap and binning artifacts, and therefore is able to reassign color to regions that would otherwise appear mixed. The algorithms can be readily applied to virtually any lambda stack generated using additive fluorescent probes, but images containing absorbing dye signatures (imaged in brightfield) and reflectance images must be converted to optical density before applying linear unmixing algorithms.

where A is the absorbance of species (i), C is the concentration, and l is the specimen optical pathlength (usually measured in micrometers for thin sections of stained tissue). The preliminary step of the calculation is to determine the optical density (OD) of each species from the measured transmission:

Commercial instruments equipped for spectral imaging and linear unmixing bundle the hardware and software into an integrated package that, in most cases, is user-friendly and highly functional. There are, however, several practical considerations that will assist in ensuring the greatest potential for success in real-world investigations. Perhaps the most important aspect is to carefully gather reference spectra for all fluorophores that faithfully represent the spectra likely to be obtained from the test specimen. In this regard, all controls must (without deviation) be imaged under identical conditions. Pixel saturation and high background noise should be avoided, and autofluorescence should be taken into consideration as a component spectrum in cases where it cannot be eliminated. In addition, unwanted signal from laser lines and mounting media discontinuities can affect linear unmixing results. Finally, low signal levels arising from dispersing the emission signal over a broad spectral range can adversely affect the results of quantitative experiments. Whenever possible, fluorophore concentrations (or at least signal levels) should be as closely matched as possible to ensure optimum results.The microscope optical train components (mirrors, lenses, filters, beamsplitters) will impose a considerable amount of bias on reference and test measurements, so collecting spectra from other instruments is unwise and will no doubt lead to errors. In order to avoid pitfalls, reference spectra should be acquired in exactly the same manner as the lambda stack, including using the same objective, pinhole diameter, offset, amplifier gain, photomultiplier voltage (not necessary when the gain responses are calibrated), wavelength range, dichromatic mirror, laser power, pixel dwell time, specimen mounting medium, coverslips, and immersion oil. The best approach is to determine the imaging parameters for the lambda stack first, and then acquire reference spectra under identical conditions. In specimens that exhibit considerable spatial separation of the fluorophores under consideration, it might be possible to chose non-overlapping regions from within the lambda stack itself for analysis. 2ff7e9595c