Magnetic Levitation Coupled with Portable Imaging and Analysis for Disease Diagnostics

Summary

We present a magnetic levitation technique coupled with automated imaging and analysis in both a smartphone-compatible device and a device with embedded imaging and processing. This is applied to measure the density distribution of cells with two demonstrated biomedical applications: sickle cell disease diagnosis and separating white and red blood cells.

Abstract

Currently, many clinical diagnostic procedures are complex, costly, inefficient, and inaccessible to a large population in the world. The requirements for specialized equipment and trained personnel require that many diagnostic tests be performed at remote, centralized clinical laboratories. Magnetic levitation is a simple yet powerful technique and can be applied to levitate cells, which are suspended in a paramagnetic solution and placed in a magnetic field, at a position determined by equilibrium between a magnetic force and a buoyancy force. Here, we present a versatile platform technology designed for point-of-care diagnostics which uses magnetic levitation coupled to microscopic imaging and automated analysis to determine the density distribution of a patient's cells as a useful diagnostic indicator. We present two platforms operating on this principle: (i) a smartphone-compatible version of the technology, where the built-in smartphone camera is used to image cells in the magnetic field and a smartphone application processes the images and to measures the density distribution of the cells and (ii) a self-contained version where a camera board is used to capture images and an embedded processing unit with attached thin-film-transistor (TFT) screen measures and displays the results. Demonstrated applications include: (i) measuring the altered distribution of a cell population with a disease phenotype compared to a healthy phenotype, which is applied to sickle cell disease diagnosis, and (ii) separation of different cell types based on their characteristic densities, which is applied to separate white blood cells from red blood cells for white blood cell cytometry. These applications, as well as future extensions of the essential density-based measurements enabled by this portable, user-friendly platform technology, will significantly enhance disease diagnostic capabilities at the point of care.

Introduction

Here, we present a technology platform and a technique which uses magnetic levitation coupled with automated imaging and analysis to analyze the density distribution of a patient's cells as an indicator for disease. This versatile approach for density-based cytometric analysis can ultimately be applied to a range of disease diagnostics. However, in order to be compatible with point-of-care testing and use in developing countries, the technique must satisfy requirements for low cost, portability, and usability. The device and consumables must be easily obtained at a low cost. The sample preparation must be simple, analysis should be automated with minimal requirements for user input or interpretation, and results should be returned quickly. Further, the device must be compact and portable to be useful in clinical settings as well as developing countries. Thus, we have developed a device and method to use magnetic levitation in point-of-care-compatible technology by coupling automated imaging and image analysis to return results regarding the density distribution of a population of a patient's cells.

Point-of-care technologies offer a notable advantage over current clinical laboratory testing procedures. The technology currently available is too expensive to be owned by a clinician or too complex to be carried out by medical staff. Many of these procedures require labor-intensive protocols which must be carried out by a trained technician. For these reasons, patient samples such as blood or urine are generally collected in the physician's office then transferred to a remote, centralized testing laboratory for clinical testing, which may take several days for the physician to receive the results of the test. This can cause delays or complications in the course of treatment in some cases, makes this testing very costly and inefficient (causing a financial burden on insurance payers), and further makes many diagnostics inaccessible in low-resource settings and developing countries.

Here, we present a magnetic levitation technique coupled with automated imaging and analysis in both a device with embedded imaging and processing (Figure 1) and a smartphone-compatible device (Figure 2). These magnetic levitation-based devices represent a broadly applicable platform technology which has the potential to be applied to a range of different medical diagnostic applications. The magnetic levitation approach functions based on an equilibrium between two forces: a magnetic force and a buoyancy force^1,2,3. When a particle is suspended in a paramagnetic medium and inserted into a magnetic field generated by two magnets with like poles facing each other, a magnetic force acts on the particle in the direction toward the centerline between the two magnets. The buoyancy force is caused by the relative density of the particle compared to the suspending medium and is upward in the case of particles less dense than the medium and downward in the case of particles denser than the surrounding medium. Based on these two forces, particles will reach an equilibrium levitation position in the field which balances these two forces; this position is directly related to the density of the particle, with denser particles levitating lower in the field than less dense particles. An imaging module, either a built-in smartphone camera^4,5,6 or independent optical components equipped with a magnifying lens^7,8, are used to visualize the positions of the particles. Image processing, either through a smartphone application^4,5,6 or an embedded processing unit^7,8, then processes the captured images to quantify the spatial distribution and, therefore, the density distribution of the population. In order to analyze larger samples (such as those with only a few particles of interest per milliliter, flow can be integrated directly into the device such the particles are levitated and analyzed as they pass through the imaging region (Figure 2).

Figure 1: Self-contained Magnetic Levitation Platform. (a) Compact magnetic levitation device including a magnetic focusing module, imaging components (a light-emitting diode (LED), an optical lens, and a camera detector), and a processing unit with a display screen. (b) Magnetic field strength in the cross-section of the area between the magnets where the sample is inserted. The field strength is greatest at the surface of the magnets and approaches zero at the centerline between them. (c) Particles, such as cells, within the magnetic field experience several forces: a magnetic force (F_m) toward the centerline between the magnetics, with magnitude varying based on the position of the particle; a gravitational force (F_g') which depends on the particle density relative to that of the suspending medium, and a drag force (F_d) resisting the particle motion. Reproduced, with permission, from Yenilmez, et al.⁸ Please click here to view a larger version of this figure.

Figure 2: Smartphone-compatible Flow-assisted Magnetic Levitation Platform. (a-c) Front (a), side (b), and back (c) views of magnetic levitation device (d) The components of the device include: 1) Magnetic levitation module, including permanent magnets, a magnifying lens, and an LED and light diffuser, 2) smartphone case, 3) electronics, including a microcontroller, pump driver, and Bluetooth receiver, 4) micro-pump holder, 5) adjustable orifice, 6) waste tube holder, 7) battery holder, 8) sample holder, 9) dual-purpose stand and cover. (e) Flow schematic, showing pumping of the sample through the magnetic field. (f) Cross-section of the magnetic levitation module, showing how particles of different densities will align as they are pumped through the field; less dense particles, such as Particle 1, will equilibrate at a higher levitation height than denser particles, such as Particle 2. Reproduced, with permission, from Amin, et al.¹ Please click here to view a larger version of this figure.

The minimum requirements for use of any sample for density distribution analysis in this system include the ability to obtain a suspension of cells or particles greater than approximately 5 µm and less than approximately 250 µm in size (for imaging and image processing) and its compatibility with mixing in a solution of a paramagnetic solution such as the gadobutrol used here. For disease diagnostics, compatible applications include those in which (i) cells of interest inherently have an altered density when they carry a disease compared to healthy controls, (ii) a density change can be induced in the cell by addition of a reagent or some alternative treatment for a short incubation time, or (iii) different cell types are being identified in a single sample and inherently (or via some treatment) have unique characteristic densities.

Sickle cell disease is a genetic disorder causing a mutated form of hemoglobin, HbS, to be produced in a person's red blood cells (RBCs), which can result in intermittent vaso-occlusive events and chronic hemolytic anemia⁹. It is diagnosed using either hemoglobin isoelectric focusing, high-performance liquid chromatography (HPLC) fractionation, or hemoglobin electrophoresis which are highly accurate but must be performed in a clinical testing laboratory because they are incompatible with point-of-care settings. Solubility and paper-based tests for sickle cell disease have been proposed, but generally require subjective user interpretation and confirmatory testing. Here, we use a density-based approach to identify sickle RBCs, which attain a higher density than RBCs from people without sickle cell disease. The mechanism involves polymerization of the mutated form of hemoglobin, HbS, which causes RBC dehydration in sickle cell disease RBCs under deoxygenated conditions^10,11,12,13.

This density-based approach can also be applied to separate cells of different types on the basis of density: white blood cells (WBCs) and RBCs⁷. WBCs are generally responsible for fighting infections in the body. WBC cytometry can be used to quantify the number of these cells in the blood and serves as a useful diagnostic tool. WBC counts higher than normal (generally considered greater than 11,000 cells per µL) may indicate infection, immune system disorders, or leukemia. WBC counts below the normal range (around 3,500 cells per µL) may be caused by autoimmune disorders or conditions which damage bone marrow. Unlike alternative technologies, the process presented here does not rely on lysis of the RBCs or stains in order to identify WBCs. This cell-based test takes advantage of the unique inherent densities of the two cell types to perform separation, as the WBC population density has been reported to be lower than that of the RBC population as calculated previously using density gradient centrifugation^1,5,8.

Compared to alterative testing at remote locations, this test is rapid, with simple sample preparation (Figure 3), separation of cells in the device within 10 - 15 min, and automated imaging and analysis which requires less than 1 min. In this way, the device can return results quickly to better inform medical decisions, allow treatment to be administered immediately to alleviate physical and psychological pain, and reduce the risk of complications associated with a delay in medical care. This technique can be performed on-site either in clinical settings due to simple sample preparation and automated imaging and analysis which returns a result with minimal user input or interpretation. Because of the use of a simple approach using permanent magnets for sample analysis and the use of either a smartphone or simple electrical components for imaging and image processing, the device as well as the per-test costs are minimal compared to some sophisticated testing procedures.

Protocol

Ethical Statement: All procedures involving human blood samples were carried out according to the institutional regulations. All protocols were reviewed and approved by the Institutional Review Board. An informed consent was given by all participants.

1. Sample Preparation for Sickle Cell Disease Diagnosis^5,8

1. Prepare a 50 mM solution of gadobutrol in Hank's Balanced Salt Solution (HBSS).
2. Dissolve 10 mM sodium metabisulfite in the gadolinium solution.
   NOTE: Sodium metabisulfite is toxic by inhalation and strongly irritates skin and tissue. It is a corrosive acid when mixed with water. It may decompose to emit toxic oxide fumes of sulfur and sodium when heated to high temperature.
3. Obtain blood sample via either fingerstick or venipuncture.
   1. Draw blood using a lancing device with a clean, disposable lancet. Apply pressure near the pierced site and wipe the blood droplet formed three to four times before collecting blood using a pipette; take care not to "milk" the finger by squeezing the tissue, as this will result in contamination of the sample with tissue fluid.
   2. Alternatively, draw blood using standard venipuncture procedures¹⁴.
      NOTE: If samples will be stored for more than a few hours, collect the blood into a vacutainer with an anticoagulant such as EDTA.
4. Add less than 1 µL of blood to 100 µl of the gadolinium-sodium metabisulfite solution.

2. Sample Preparation for WBC Cytometry⁷

1. Obtain blood sample via either fingerstick or venipuncture, as in section 1.
2. Optional: Lyse RBCs with an RBC lysis buffer. Pipette 5 µL of blood into 500 µL of RBC lysis buffer and incubate at room temperature for 3 - 5 min.
   NOTE: this may be done to confirm the levitation range of an isolated population of WBCs. However, it is not necessary to perform this step, as the results presented here demonstrate that RBCs levitate at a distinctly lower position than WBCs.
3. Dilute whole blood 1:1,000 in 25 mM Gd in HBSS
   NOTE: if cell lysis was performed, dilute the lysed sample 9:1 lysed sample:250 mM Gd to obtain a sample containing 25 mM Gd.

3. Analysis of Samples Using the Magnetic Levitation Platform^4,5,6,7,8

1. Start the magnetic levitation device:
   1. For the self-contained device, plug in the device, allow it to power up, and place on a flat, level surface.
   2. For the smartphone-compatible version, launch the smartphone application and swipe left to enter image capture mode, and place on a flat, level surface.
2. For static magnetic levitation:
   1. Load the prepared sample into a square glass microcapillary tube by dipping the end into the solution and allowing the sample to fill the capillary via capillary action.
   2. Seal the end with a tube sealant by slowly pushing one end the capillary into the material.
   3. Insert the capillary between the magnets of the magnetic levitation device so that only 1 cm of the tube remains visible (please see Figure 3 for an illustration the sample preparation) .
      NOTE: This step remains the same for both the smartphone-compatible and self-contained device.
   4. Wait 10 min without disturbing the device or the capillary.
3. For flow-assisted magnetic levitation:
   1. Load the sample into the sample tube.
   2. Connect the inlet tubing between the sample container and microcapillary and connect the outlet tubing between the microcapillary and waste container.
   3. Set the LED intensity and flow parameters.
4. Ensure that cells are visible in the field of view, then press the capture button to capture an image (button 3 on the self-contained device and the camera button on the bottom of the screen in the smartphone application).
   NOTE: If a USB is used to store the images, create a folder called "images" and insert the USB prior to turning on the device. If no drive is present, the device will store the images in its internal memory and transferred later.
   NOTE: Several images may be captured and analyzed to reduce the risk of anomalies by moving the capillary in or out about ½ cm between image capturing.
   NOTE: If the number of cells within the field of view is too high or too low (as detected and reported on the user interface), shift the capillary in or out of the device or prepare another sample.
5. Remove the sample and discard the sample according to the institutional or local regulations. Capillaries should be disposed as a sharp.

Figure 3: Sample Preparation and User Interface. (a) The sample preparation procedure involves lancing the subject's finger, forming a droplet of blood, transferring the blood drop to the sample testing solution, agitating the sample and loading into a capillary tube via capillary action, and inserting the sample into the magnetic levitation device. (b) These sample preparation steps are also displayed on the screen of the device to guide sample preparation. (c) The device includes four buttons: a button to zoom into the sample image in order to properly adjust the focus using the adjustment knob; a button to obtain a single measurement (a 5 s delay is implemented to allow time for the user to insert the sample); a time-lapse measurement (6 images are taken at 5 s intervals); and a button to power off the device after use. Reproduced, with permission, from Yenilmez, et al. ⁸Please click here to view a larger version of this figure.

4. Image Analysis^4,5,6,7,8

1. Run the image analysis software included on the device for the appropriate sample (cell distribution or cell type separation).
   NOTE: For the self-contained device, the analysis is performed automatically and displayed on the graphical user interface. For the smartphone-compatible device, to perform the analysis, go to the gallery and select the desired video file to be analyzed.
2. Observe and record the output of the analysis displayed on-screen.
   NOTE: For cell distribution analysis (such as sickle cell disease diagnosis), the output will be the width of confinement of the cell population.
   NOTE: For cell type separation (such as WBC identification), the output will be an image with the WBC population identified. In order to calculate the number of cells per microliter, multiply the average number of WBCs per image by a factor of 2,000. For example, if 5 WBCs were observed, this would indicate 10,000 WBCs/µL. The normal range is generally considered to be 3,500 - 11,000 WBCs/µL.
3. Repeat the analysis by moving the sample tube in or out of the device about ½ cm and capturing another image to be analyzed as described above.
   NOTE: It is recommended to repeat the analysis 5 - 6 times per sample to avoid errors.

Results

For cell density distribution analysis, which is the technique used for sickle cell disease diagnosis, the aim is to identify the width of the distribution of the cell population. Blood cells from patients without sickle cell disease will be confined within a predictable width. Cells from patients with sickle cell disease will be distributed throughout a wider region, with a downward skew in the cell distribution (see Figure 4.) For any particular application, a threshold...

Discussion

Critical Steps within the Protocol
Critical factors in this process include the proper alignment of the magnets. If the magnets become dislodged or separated more than normal within the device, this can affect the results. To control for this fault or others in the process, a density-controlled particle, such as polystyrene microspheres, can be used periodically to control for changes over time. Further, levitation time is important to allow the cells to reach equilibrium. For red blood cells, 10 min is sufficient ...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Gadavist (Bayer)	Jefferson Medical and Imaging	2068062	Gadavist contains 1M gadobutrol, a chelate of gadolinium. We purchased 2 ml vials with 15/ca.
Square glass microcapillary tubes	Vitrocom	8270	50 mm length is sufficient
Sodium metabisulfite	Sigma-Aldrich	S9000	Chemical formula: Na2S2O5
Leica Microsystems Critoseal tube sealant	Fisher Scientific	02-676-20
Hank's Balanced Salt Solution	Sigma-Aldrich	H9269 SIGMA
Trypsin-EDTA	Sigma-Aldrich	T4049	Or other reagent as recommended for the cell type used
MICROLET 2 Adjustable Lancing Device	Walgreens	246567	Any lancing device is acceptable when used according to biosafety protocols
Microlet Lancets	Walgreens	667474	Must be dispoable and not reused
Hausser Bright-Line Phase Hemacytometer	Fisher Scientific	02-671-6	Or any preferred method for cell counting
ACK Lysing Buffer	ThermoFisher	A1049201