Quantitative imaging with a mobile phone microscope

Abstract

Use of optical imaging for medical and scientific applications requires accurate quantification of features such as object size, color, and brightness. High pixel density cameras available on modern mobile phones have made photography simple and convenient for consumer applications; however, the camera hardware and software that enables this simplicity can present a barrier to accurate quantification of image data. This issue is exacerbated by automated settings, proprietary image processing algorithms, rapid phone evolution, and the diversity of manufacturers. If mobile phone cameras are to live up to their potential to increase access to healthcare in low-resource settings, limitations of mobile phone-based imaging must be fully understood and addressed with procedures that minimize their effects on image quantification. Here we focus on microscopic optical imaging using a custom mobile phone microscope that is compatible with phones from multiple manufacturers. We demonstrate that quantitative microscopy with micron-scale spatial resolution can be carried out with multiple phones and that image linearity, distortion, and color can be corrected as needed. Using all versions of the iPhone and a selection of Android phones released between 2007 and 2012, we show that phones with greater than 5 MP are capable of nearly diffraction-limited resolution over a broad range of magnifications, including those relevant for single cell imaging. We find that automatic focus, exposure, and color gain standard on mobile phones can degrade image resolution and reduce accuracy of color capture if uncorrected, and we devise procedures to avoid these barriers to quantitative imaging. By accommodating the differences between mobile phone cameras and the scientific cameras, mobile phone microscopes can be reliably used to increase access to quantitative imaging for a variety of medical and scientific applications.

Figure 1. A multi-phone mobile microscope.

A Diagram of the magnifying optics and illumination added to a mobile phone to create a transmission light microscope. B Prototype of a field-ready mobile microscope – the CellScope – that has a folded optical path for compactness and is equipped with a multi-phone holder and iPhone 4. Phone-specific variants have been evaluated on five continents for various applications. C A Wright stained blood smear taken on the mobile microscope with an iPhone 4 and 20×/0.4 NA objective showing the inscribed field of view captured by the device. D Enlarged images of the small region of interest in C containing a granulocyte and red blood cells taken with four different mobile phones. The images demonstrate resolution, color, and brightness differences among phones.

Figure 2. Spatial resolution of mobile phone microscopy is dependent on microscope optics.

A The resolution that can be captured with a mobile phone microscope approaches that of a scientific camera coupled to the same optics across a range of numerical apertures. Inset shows the measured intensity profile across bars of non-transmitting chrome spaced at 512 line pairs per millimeter and taken with a 10×/0.25 NA objective, as well as the ideal target profile. The Michelson contrast calculated for this example group is 41%, indicating that features with this spacing are resolved. B Wright stained blood smear with an inset of a granulocyte and red blood cells taken with a 10×/0.25 NA objective and iPhone 4. C Image of the same sample and region of interest taken with a 40×/0.65 NA objective and iPhone 4 showing improved resolution.

Figure 3. Illumination variation, image distortion, and pixel non-linearity of mobile phone microscopy can be minimized.

A Variation in illumination across a clear field of view along the horizontal and vertical axes for an LED flashlight source evaluated with a 10×/0.25 NA objective and iPhone 4. B Distortion across a field of view, evaluated along a bar target with a 10×/0.25 NA objective and iPhone 4. A parabolic fit is superimposed on the data for both axes. C The measured pixel response (+) of an iPhone 4 and 10×/0.25 NA objective to changes in illumination intensity is nonlinear but can be corrected for the gamma encoding (o) to recover a linear pixel response (R² = 0.999).

Figure 4. Spatial resolution of mobile phone microscopy has improved with mobile phone advancement.

A The spatial resolution of mobile phone microscopy with iPhone and Android phones is plotted as a function of the effective pixel size for images taken with a 10×/0.25 NA objective. The theoretical constraints on resolution imposed by pixel spacing on the Bayer color sensor array are plotted along with the empirically determined resolution limit of the underlying microscope optics. B The spatial resolution of mobile phone microscopy with the same iPhone and Android phones is plotted as a function of megapixel count. C Spatial resolution of the iPhone family of phones is plotted over time, together with the dates of significant camera advancements.

Figure 5. Mobile phones differ from scientific cameras in selection of image capture and processing parameters.

A Common core hardware components underlie the capture process of both mobile phone cameras and scientific cameras. B The capture and processing parameters are set directly through the user interface of a scientific camera. C On mobile phones, an intermediate layer assesses the view of the camera in real-time and modifies image acquisition. This simplifies the user interface for traditional point-and-shoot photography but sacrifices the control desired by a scientific user.

Figure 6. Phone automation during image capture and storage degrades feature and color information.

A High-contrast image of an array of 20 µm diameter pinholes in chrome taken with an iPhone 4 and 20×/0.40 NA objective. B Phone auto-focus changes effective magnification for samples held at different distances from the objective lens, resulting in changes to apparent feature size. C Saturation in individual color channels due to auto-exposure and gain causes loss of color contrast in sparse, bright samples. D Built-in white-balance of the iPhone 4 is insufficient to overcome variable illumination conditions, resulting in a shift of the apparent color profile of samples such as this blood smear, as quantified in E and F for a region of interest illuminated with a white LED or halogen lamp, respectively. G Sharpening algorithms cause patterns of artificial ringing and haloing around high contrast structures such as this chrome-on-glass resolution target. H Intensity profiles at edges show phone-specific deviations from the expected theoretical monotonic profile obtained from an incoherently illuminated sample using a scientific camera and no image processing. The measured profiles are normalized, with 0 corresponding to the intensity in chrome far from an edge and 1 corresponding to the intensity in a clear region far from an edge.

Figure 7. Proposed steps to enable quantitative, reproducible imaging with a mobile phone microscope.

A Standardize illumination source and brightness. B Set focal state on a field with known dimensions or features. C Set exposure and gain using a clear field of view. Use this field to set or select a white balance state; may require resetting exposure and gain to ensure changes in white balance do not result in saturation of a color channel. D Acquire images of samples while keeping capture settings constant. E Information content can be preserved by selecting lossless or high quality compression settings. In addition, multiple images can be used to record additional z planes or expand the effective dynamic range of the image. While many of the features required to implement these steps are not directly accessible in the default camera, they are built into commonly available third-party camera applications or can be incorporated into custom applications.