1 Introduction
COVID-19 is a human infectious disease caused by a novel coronavirus, which has posed a great threat to public health in more than 200 countries and regions around the world. As of mid-March 2021, the cumulative number of confirmed cases worldwide exceeded 120 million, resulting in more than 2.7 million deaths. Therefore, it is very important to find a rapid and accurate diagnostic method to contain COVID-19, which can help to detect and control the source of infection in time to avoid the spread of the epidemic. Traditional detection methods are laboratory-based, usually equipped with heavy analytical equipment to work, and require professional operators, which hinders the efficient prevention and control of COVID-19.Especially in remote areas with limited resources, problems such as backward medical technology, underdeveloped transportation and weak awareness of health and safety have laid hidden dangers for the outbreak of the epidemic
[1]. Smartphone-based point-of-care detection technology (POCT) is an effective solution to the above problems, which achieves more economical, accurate and rapid detection than traditional methods by combining micro-sensors with smartphones. At present, this method has a wide range of applications in health care, precision agriculture, environmental protection, food safety, military protection and other fields
[2,3][4⇓⇓~7][8⇓⇓~11][12,13][14].
Smartphone-based POCT has the following unique advantages: (1) it is easy to operate and does not require professional training; (2) The test result can be obtained in time without waiting for a long time; And (3) that manufacture cost is low, and the method is beneficial to use in areas with limit resources. These characteristics are helpful for people to make rapid decisions under non-ideal conditions, which is of great value in the field of medical and health care and represents the development direction of this emerging technology. In recent years, many researchers have introduced the current applications of POCT based on smart phones and discussed its development prospects, most of which have proposed solutions to various human health problems
[15⇓⇓⇓~19]. Body fluid is a kind of sample that can be directly collected from the human body, which is of great significance for the diagnosis of diseases, especially in resource-poor areas, because human tissue fluid or metabolites can be directly used as test samples to reduce the cost of reagent preparation and facilitate users. Breathing gases are also of great value, especially volatile organic compounds (VOCs), which are ubiquitous in the environment. The detection of inhaled VOCs can help people avoid exposure to toxic environments, while the sensing of biomarkers in exhaled VOCs can help people better understand their health status. In addition, POCT is often used in the detection of vital signs, such as heart rate, blood pressure, body temperature, pulse and so on. Therefore, in this review, we use the objects detected by POCT (body fluids, VOCs, vital signs) as the basis for classification, and combine with the current mainstream sensing strategies to systematically review the practical application of smartphone-based POCT in the past three years. Finally, the development needs and future prospects of the field are summarized and discussed.
2 Body fluid detection
Because it does not require expensive reagents and large medical equipment, POCT based on body fluid detection is particularly suitable for daily production and life. At present, the body fluids that can be used for sensing analysis mainly include blood, sweat, tears, saliva, urine, etc. By effectively extracting and processing these body fluids, the required physiological characteristics information can be obtained.
2.1 Blood
Serum and plasma in blood, are widely used as samples in routine clinical analysis. Blood contains many components closely related to the physiological state of the body, such as proteins, nucleic acids, etc. Accurate blood detection through POCT devices based on smart phones is a major research hotspot at present
[20,21].
2.1.1 Colorimetric sensing
Colorimetry is a commonly used biochemical determination method in sensing analysis, which is mainly detected by evaluating the change of absorbance or reflectance of samples, while nanotechnology provides many materials with excellent optical properties for colorimetric analysis
[22⇓⇓⇓~26][27,28]. Among them, spherical nucleic acids based on gold nanoparticles are widely used in the field of biosensing due to their good stability and easy observation. Inspired by this progress, Liu et al. Proposed a smartphone-based colorimetric detection method for protease biomarkers through the modular combination of proteolysis reaction transcription and spherical nucleic acid for colorimetric analysis of protease biomarkers (Fig. 1A)
[29]. Proteolysis directed by the target protease generates an active RNA polymerase, which is then transcribed in vitro to produce multiple RNAs with programmable sequences. Transcribed RNA triggers the assembly of spherical nucleic acids through base pairing, which leads to a significant red shift in the absorption wavelength associated with SPR and a significant change in the color of the test paper. The assay achieved specific detection of matrix metalloproteinase-2 (MMP-2) with a lower limit of detection of 3.3 pM. The modular combination of proteolysis reaction transcription and spherical nucleic acid has strong signal amplification ability and realizes high-sensitivity colorimetric detection. In addition, this colorimetric method has good universality, and can detect other protease biomarkers such as thrombin and hepatitis C virus by adjusting the target RNA polymerase module.
2.1.2 Fluorescence sensing
Fluorescence analysis is widely used in biomedical research because of its high sensitivity, low reagent consumption and strong specificity
[30⇓~32]. Zhang et al. Developed a mobile phone-based acoustofluidic platform that can be used to detect hemoglobin in human blood (Fig. 1B)
[33]. The platform uses acoustic radiation force to enrich nanoscale analytes and red-green fluorescent nanoparticles as fluorescent probes for POCT visual detection, so that the color signal from the probe is enhanced and the sensitivity of colorimetric sensing is significantly improved. This approach, which combines a voice-activated nanoparticle platform with a mobile phone-based fluorescence assay, introduces a new development path for POCT.
2.1.3 Bioluminescence sensing
Sensing proteins based on bioluminescence resonance energy transfer (BRET) are considered ideal reagents for POCT because their ratiometric signals are readily detectable. Tomimuro et al. Developed a microfluidic wire-based assay device (μTADs) as a new platform for antibody detection using BRET-sensing proteins
[34]. The luminescent antibody sensor protein and luminescent matrix were pre-deposited on two tangled cotton threads in dry form (Fig. 1C), and after responding to specific antibodies, rapid bioluminescence detection was performed through a mobile phone camera (Fig. 1D). The microfluidic technology used in the device is an emerging technology involving chemistry, fluid physics, microelectronics, biology and other disciplines. Due to the difference in the performance of fluids at the micro scale and the macro scale, the use of micron-scale pipes to manipulate fluids can achieve a series of processing and treatment that are difficult to achieve by conventional methods. μTADs are inexpensive to manufacture and can be used for blood testing in small sample sizes, and the natural fabric properties of cotton thread make them more flexible in design.
2.2 Sweat
Sweat is an important metabolite, containing uric acid, lactic acid and various electrolytes. Sweat detection has been favored because of its non-invasive nature, but how to collect samples conveniently has been puzzling researchers. The emergence of flexible electronics gives people hope to use ductile substrates to make electronic devices more flexible, to a certain extent, to meet the deformation needs of equipment, and to promote the development of wearable sensors
[35⇓⇓⇓~39]. On the basis of flexible electronic devices, researchers have developed many portable sensors for sweat detection, which have made outstanding contributions to the personalization and intelligence of health care
[40,41].
2.2.1 Electrochemical sensing
Electrochemical sensor is a kind of sensor that uses the electrochemical properties of the substance to be detected to convert its chemical quantity into electrical signals for sensing and detection. POCT devices based on electrochemical sensing are widely used in clinical diagnosis, industrial analysis and environmental assessment due to their advantages of accuracy, rapidity and low cost
[42]. Yoon et al. Introduced a wearable sweat sensor made of a robust bio-based supramolecular polymer (Fig. 2A)
[43]. A linear potentiometric ion sensor was fabricated by coating a carbon fiber wire electrode with a citrate-based supramolecular polymer. These flexible sensors can detect both K
+ and Na
+ and can be woven into textiles to create a wearable multi-target ion sensor. Moreover, the sensor is capable of self-healing via hydrogen bonding, showing a self-healing efficiency higher than 97% after 30 s of rupture at room temperature. The wireless transmission circuit can send data to the smartphone user interface and monitor electrolyte ions in human sweat in real time even during self-healing.
2.2.2 Colorimetric sensing
Detection diversification is a major development direction of sensors. Because of the rich composition of sweat, how to analyze multiple samples at the same time has been a hot topic in the field of sweat detection. Xiao et al. Reported a wearable colorimetric sensing device based on a microfluidic chip for the detection of glucose in sweat
[44]. The device has five microfluidic channels emanating from the center (Fig. 2B), which can direct sweat from the epidermis to the detection microchambers, each of which has a check valve (Fig. 2C) to avoid backflow of chemical reagents. The microchamber contains a pre-embedded sensing reagent that reacts with glucose in sweat to produce a color change. The sensor can perform five parallel detections at the same time, resulting in five times the detection efficiency of a conventional sensor. Bandodkar et al. Designed a wearable sweat sensing device that can perform both colorimetric and electrochemical analysis
[45]. The device combines a chronometric microfluidic platform with an embedded colorimetric assay (Figure 2D) and is powered by a biofuel cell that simultaneously monitors sweat rate, pH, lactate, glucose, and chloride. Compared with sensors that detect a single parameter, the device can provide a more comprehensive human health information.
2.2.3 Fluorescence sensing
Sekine et al. Integrated the fluorescence sensing mode onto a flexible microfluidic device and proposed a sensing device for in situ quantitative analysis of sweat chemistry (Fig. 2 E, F)
[46]. The microfluidic device consists of a microchannel network array and microreservoirs pre-filled with fluorescent probes that selectively react with target analytes (chlorine, sodium, and zinc) in sweat.Resulting in a change in fluorescence intensity, which is detected by a smartphone equipped with an optical module, resulting in a quantitative and rapid analysis. The results of human trials show that the device can accurately measure the concentration of target biomarkers in sweat with an accuracy comparable to that of traditional laboratory technology.
2.3 Saliva
Saliva is a complex mixture, mainly secreted by salivary glands. It is colorless, tasteless and acidic. The components of saliva include various enzymes, electrolytes, proteins, nucleic acids, etc. Among them, there are more than 1000 kinds of proteins, most of which have detection value. Compared with blood collection, saliva collection is safer and more convenient, and because it is non-invasive, there is no risk of disease transmission. At the same time, saliva collection has the advantage of real-time sampling compared with sweat collection. Therefore, in recent years, researchers have proposed many POCT schemes for saliva detection, including many studies of great significance for the prevention and control of the COVID-19 epidemic
[47⇓~49].
2.3.1 Electrochemical sensing
In order to facilitate the collection of saliva, Chen et al. Designed a smart toothbrush, which can be used to monitor the glucose in saliva in real time after matching with a smart phone
[50]. Almost all glucose sensors currently available on the market use glucose oxidase or dehydrogenase for detection. However, the activity of enzyme is easily affected by external conditions such as temperature, pH, ion concentration and so on, which inhibits the sensitivity of sensing. Therefore, glucose detection based on non-enzymatic method has been pursued. Chen et al. Found that tin bronze is a suitable non-enzymatic material for salivary glucose sensing, and compared with other types of bronze, tin bronze has the advantages of being stronger and easier to cast. An electrochemical sensor using a bare copper electrode as the working electrode is integrated into a smart toothbrush to monitor salivary glucose, which has potential application value for non-invasive detection of diabetes. In addition, since no additional reagents are involved in the detection, the electrode can be properly stored after simple mechanical polishing for long-term use.
2.3.2 Fluorescence sensing
In today's epidemic, it is necessary to develop a POCT device that can detect COVID-19. At present, most nucleic acid tests require nasopharyngeal or oropharyngeal swab samples, which must be collected by professional medical personnel wearing complete protective equipment, and the collection process is very painful and unbearable. In order to improve the current situation, Ning et al. Developed a portable COVID-19 detection device based on saliva analysis (Fig. 3A), which takes only 15 min from sampling to obtaining results
[51]. The protocol is very simple (Figure 3B). Saliva samples are first collected in a tube pre-filled with 3 mL of lysis buffer and then heated for about 5 minutes at a temperature above 37 ℃. Following this, 5 μL of lysis sample was added to each reaction chamber containing 10 μL of premixed RPA and CRISPR solution on the detection chip. Finally, the chip was incubated at room temperature for 10 min, inserted into the smartphone reader, turned on the laser diode, and the results were captured by the smartphone camera. Saliva samples have the advantages of convenient collection, non-invasive, and they can be obtained directly by the patient, without special materials, professional training or perfect infrastructure, reducing the involvement of medical staff and eliminating the risk of exposure. The COVID-19 testing protocol developed based on these characteristics has a far-reaching impact in breaking through the current testing bottleneck and assisting the epidemic screening efforts.
2.4 Tear
Tears are a complex fluid composed of lipids, electrolytes, proteins, peptides, glucose and amino acids, which contain a variety of biomarkers, such as immunoglobulin, albumin, Na
+, K
+ and Cl
-. In addition, many components in tears are similar to those in blood. For example, tears were once widely used to monitor blood glucose levels in diabetic patients as a substitute for blood because of their high correlation with capillary blood glucose. Therefore, smartphone-based tear detection has attracted great attention due to non-invasive, no pretreatment, low potential interference and direct correlation with biological fluids such as blood. However, compared with blood, saliva and other body fluids, it is difficult to obtain tears. It is of great significance to design an intelligent device that can collect, detect and analyze tears in time for POCT based on tears. In recent years, wearable and implantable intraocular pressure biosensors have attracted much attention. By tracking the fluctuation of intraocular pressure (IOP), they can convert the tiny IOP pressure signal into easily observed inductive, resistive or optical signals, and then accurately detect diabetes, liver diseases and ophthalmic diseases
[52].
2.4.1 Colorimetric sensing
Smart contact lenses have gained a lot of attention due to their ability to directly monitor the physiological information of tears, but the current promotion is still limited, because the successful development of smart contact lenses needs to overcome many problems such as mechanical biocompatibility, detection sensitivity and challenging integration
[53]. In addition, they should have a high visual field transparency and be comfortable to wear to minimize the possibility of inflammation caused by eye irritation. Moreddu et al. Developed a microfluidic contact lens that can detect pH, glucose, protein and nitrite ions in tears
[54]. The microfluidic channel consists of a central ring with four branches, which is engraved on the contact lens by laser etching technology, and the biosensor is embedded in the microcavity at the branch end. To achieve accurate reading results, smartphone imaging techniques were incorporated and further processed using a MATLAB algorithm based on the nearest-neighbor method. The results of artificial tear test showed that the sensing device could complete the response within 15 seconds. Also based on the form of contact lenses, Moreddu et al. Also developed a multiplex detection device for clinical analysis in combination with paper-based microfluidics, including pH, protein, glucose, nitric acid and ascorbic acid (Fig. 4), which can be directly sampled from tears through contact lenses
[55]. Contact lenses are made of poly (hydroxyethyl methacrylate) and sealed by chemical bonding. Microfluidic channels carved by laser etching technology can minimize leakage. A smartphone-based readout method was employed to collect and store the information, enabling a more intuitive presentation of the detection data of the analytes. Smart contact lenses have shown potential in medical diagnosis and disease screening, and will be useful in monitoring eye infections, uveitis, diabetes, keratosis and evaluating oxidative stress in the future.
2.4.2 Image detection
Visual image detection is to convert the detected object into an image signal and identify it according to the information of pixel distribution, brightness and color. Lin et al. Developed a sensor with swelling and contraction functions in the form of contact lenses, which can detect glucose in tears (Fig. 5)
[56]. The device is made of hydroxyethyl methacrylate based on phenylboronic acid, a material that swells when it absorbs glucose, a property that allows contact lenses to change thickness depending on the concentration of glucose. After the light-emitting diode device emits light, the change of glasses thickness can be detected by smartphone and analyzed by software to get the information of glucose concentration. The results of this study are of great help for non-invasive measurement of glucose levels and simple glucose sensing using smartphones.
2.5 Urine
Urine is a liquid excretion discharged from the body through the urinary system in order to meet the needs of metabolism. Its components include urine hemoglobin, enzymes, electrolytes and so on. In the medical field, urine test can help doctors diagnose many diseases and is a commonly used sample. In order to further develop the advantages of non-invasive and easy access of urine analysis, many POCT protocols for urine testing have emerged in recent years
[57⇓~59].
2.5.1 Colorimetric sensing
He et al. Introduced a biosensor capable of multi-channel rapid urine testing by combining a tape-based sensor with a diaper
[60]. Through the physical punching and fabric filling process, the micro-sensor array was arranged on the flexible waterproof tape and then adhered to the diaper to form a wearable urine sensing system. Compared with traditional urine detection, one of the advantages of this device is that it can directly sample the excreted urine through the diaper and induce the color reaction of the biosensor on the diaper without any additional operation. The smart phone is used to collect the colorimetric signal, and the goal of rapid detection is achieved. This device can detect urine markers such as glucose, nitrite, protein and phenylpyruvic acid, which is of great significance for early warning of some diseases, and is a very suitable POCT protocol for infants and patients with urinary incontinence.
2.5.2 Fluorescence sensing
The concentration of uric acid in urine is closely related to cardiovascular and cerebrovascular diseases. Developing a rapid and efficient method to detect uric acid can prevent the outbreak of diseases to a certain extent and reduce its harm to the human body. Lu et al. Proposed a POCT method for hyperuricemic patients, in which uric acid concentration was evaluated by the fluorescence detection effect of H
2O
2 after enzymatic reaction (Fig. 6)
[61]. The sensor in the method is an alginate hydrogel microsphere embedded with Cd, Zn, Te and S quantum dots and urate oxidase, and has the advantages of improving the fluorescence stability of the quantum dots, reducing protein interference, improving the response speed and the like. Through the smartphone camera and the supporting application software, the test results of uric acid concentration can be obtained within 10 minutes. Experiments using artificial urine samples have shown an accuracy of up to 100%. In addition, the method is universal, and the alginate hydrogel microsphere embedded with different enzymes can be used for detecting other diseases, showing a good application prospect.
Alves et al. First reported a fluorescence immunoassay method for the detection of Escherichia coli in urine. The microfluid used in this method has ten pores and a relatively large surface area, which enables a large number of samples to pass through the microcapsule coated with capture antibody, thus improving the detection limit
[62]. With the help of a smartphone camera, it takes only 25 minutes to detect and quantify E. coli in synthetic urine. This convenient and inexpensive analytical method is helpful to prevent and treat infectious diseases caused by Escherichia coli, especially in areas with poor environmental water quality and difficulties in sanitation and epidemic prevention.
3 Detection of volatile organic compounds
At present, respiratory gas-based POCT devices are mainly used for the detection of volatile organic compounds (VOCs). According to the definition of the World Health Organization, VOCs are a general term for volatile organic compounds with melting points below room temperature and boiling points between 50 and 260 degrees Celsius, including aldehydes, ketones, alkanes and hydrocarbons, which widely exist in indoor and outdoor environments. In fact, most VOCs are closely related to environmental assessment and disease diagnosis. For example, formaldehyde is a common irritant gas that affects human health, especially in newly decorated homes, the concentration of formaldehyde often exceeds the standard, and inhalation of excessive formaldehyde can lead to systemic diseases with system damage as the main cause. The sensing methods of VOCs include optical, chemiresistive and electrochemical techniques, which face the problems of reliability, accuracy and selectivity in continuous monitoring, and most of these problems are due to the difficulty of maintaining high stability of sensors in complex environments. Therefore, in order to overcome the selectivity barrier, sensor arrays have been developed in the direction of multimodal sensing, and the improvement of reliability and accuracy has been achieved by the progress of nanotechnology. The use of various forms of nanomaterials not only enhances the sensing performance, but also plays an important role in miniaturized detection
[63].
3.1 Hydrocarbon gas
Hydrocarbon gases from transportation and petrochemical products are the most common toxic VOCs in indoor and outdoor environments. The detection of these compounds is important for early warning of air pollution and health threats, and helps to identify the source of pollutants.
3.1.1 Image detection
The standard detection method of hydrocarbon gas is gas chromatography-mass spectrometry, but this method requires bulky, expensive equipment, and professional personnel to operate and maintain, so it is not convenient to apply in many occasions. Photoionization detector is a widely used portable VOCs detector, which can provide the detection of hydrocarbons with low selectivity. Its cost is much lower than that of gas chromatography combined with mass spectrometry, but it still costs thousands of dollars. Metal oxide semiconductor sensors, although relatively cheap and portable, are limited by poor selectivity and stability. In view of the above problems, Qin et al. Developed a hydrocarbon gas sensor with simple operation and low cost (Fig. 7), which uses the principle of cantilever bending caused by polymer swelling paper to respond and read the displacement of the free end of the cantilever through the camera of a smart phone, and then detects the concentration of hydrocarbons in the air
[64]. The sensor shows good linear response, wide detection range and fast response speed, which can meet the requirements of hydrocarbon compound detection in oil industry safety or air quality monitoring, and also has certain significance for personal health protection.
3.1.2 Colorimetric sensing
Photonic crystal sensor is a new intelligent sensor in VOCs detection, which has good visual recognition effect and reaction speed. Kou et al. Designed a photonic crystal colorimetric sensor for the detection of benzene, toluene, xylene and 1,2,4-trimethylbenzene by alternating combination of polystyrene-acrylic acid and titanium dioxide nanoparticles
[65]. Among them, polystyrene is an aryl polymer, which has high sensitivity to aromatic hydrocarbons and is used as a low refractive index material. Titanium dioxide can provide a stable photonic crystal framework and form a large refractive index contrast with polystyrene, so it is selected as a high refractive index material. The two were assembled on a silicon wafer by a spin coater to prepare a photonic crystal sensor. When exposed to a stream of toluene, the sensor rapidly produced a red shift of more than 500 nm within 1.5 s. The concentration value can be easily obtained by using the colorimetric analysis program developed on the smartphone for color capture. The sensor can be used as a high-performance aromatic VOCs sensor for environmental quality monitoring.
3.2 Formaldehyde
Formaldehyde is one of the most dangerous air pollutants. In 2017, the International Agency for Research on Cancer of the World Health Organization listed formaldehyde as a carcinogen. In daily life, excessive formaldehyde concentration in newly decorated houses is a common problem, which seriously threatens the health of residents. Therefore, it is very critical to develop a low-cost and simple method for the accurate determination of formaldehyde.
3.2.1 Colorimetric sensing
Guo et al. Designed a colorimetric sensor (Fig. 8) using the shooting function of a smartphone for the determination of formaldehyde gas
[66]. The sensor is based on the method of 4-amino-3-hydrazino-5-mercapto-1,2,4-triazole for the determination of formaldehyde, and the sensor chip consists of two reagent libraries, a reaction library and a mixed column. At the same time, to prevent the fluid from flowing out when letting in gas molecules, a hydrophobic porous polytetrafluoroethylene membrane was placed on top of the reaction reservoir. The experimental results show that the sensor has high sensitivity and selectivity.
3.2.2 Electrochemical sensing
Formaldehyde in the air is closely related to asthma attacks, especially for children, excessive formaldehyde can easily lead to the deterioration of pediatric asthma. To solve this problem, Li and others developed a watch based on electrochemical fuel cell and Internet of Things technology, which can be used to monitor formaldehyde concentration in real time and upload the data to mobile phones through Bluetooth to help users detect potential formaldehyde threats in time
[67]. The sensor in the watch is a two-electrode electrochemical sensor, where formaldehyde in the air reacts on the surface of the electrode to convert chemical energy into an electrical signal that is easy to detect. In addition, cloud technology is used to store and analyze sensor data. The results of this study realize the visualization of formaldehyde monitoring, which can continuously and effectively monitor the formaldehyde exposure level of asthma patients and help to achieve personalized asthma treatment.
3.3 Acetone and ethanol
Exhaled VOCs from the human body can be used as biomarkers for noninvasive identification of disease for diagnosis in the field, allowing timely understanding of healthy and diseased metabolic status. Studies have shown a clear link between acetone concentration in the breath and blood glucose levels. Therefore, diabetes mellitus is often detected non-invasively by exhaled acetone in clinical medicine. Ethanol is commonly known as alcohol. The main organ of ethanol metabolism in the human body is the liver. Long-term intake of ethanol may cause alcoholic fatty liver, alcoholic hepatitis, cirrhosis and other diseases. The detection of ethanol in respiratory gas can effectively provide early warning of prevention and treatment.
3.3.1 Electrochemical sensing
Shrestha et al. Proposed a method for measuring blood glucose levels using VOCs such as acetone and ethanol, which uses two chemical sensors based on metal oxide semiconductors and can detect acetone and ethanol at a concentration of one part per million
[68]. In real-time testing, the accuracy of classifying low and high blood glucose concentrations was as high as 97%. The two sensors are integrated into an Arduino controller platform-based smart wristband (Fig. 9a), which is powered by a 3.7 V, 500 mAh rechargeable lithium-ion battery. The smartphone gets the data information of the wristband through Bluetooth and uses cloud technology to save the data.
3.3.2 Ultrasonic sensing
Ultrasonic sensor is a kind of sensor that converts ultrasonic signals into other energy signals (such as electrical signals). It has the advantages of high measurement accuracy, stable operation and good directivity, and is widely used in biomedicine, industry, national defense and other aspects. Yoon et al. Developed a portable sensor with three capacitive ultrasonic sensor arrays (CMUTs) (Figure 9C), which uses a readout system to detect a variety of VOCs (acetone, ethanol, toluene, etc.)
[69]. In order to achieve selective detection, the three sensor arrays were coated with phenyl peptide, single-walled carbon nanotubes, and a colloid of phenyl peptide, polystyrene-co-allyl alcohol, respectively. At the time of detection, when VOCs are absorbed on the sensor, the mass of the resonant structure increases, causing the resonant frequency to decrease. By measuring the shift in the resonant frequency, the concentration of VOCs can be estimated (Figure 9B). Using Bluetooth to transmit data, the result can be notified to the user at the first time. The advantages of multiple detection of the device make it a promising portable VOCs monitoring solution.
4 Vital sign detection
Vital signs are important information used to measure the severity of a patient's condition, including body temperature, heart rate, respiration, pulse, blood pressure, blood oxygen, pupils, etc. The changes of these indicators are related to the safety of patients, so it is very important to monitor the vital signs continuously and in real time. However, due to limited medical resources, not every patient can receive perfect monitoring, and POCT based on smart phones provides an effective solution to this problem
[70,71].
4.1 Pulse and blood pressure
At rest, the pulse rate of normal people is 60 to 100 beats per minute. When cardiac insufficiency, shock, high fever, anemia, pain and drug poisoning occur, the pulse rate will change significantly. Real-time monitoring of pulse information is of great significance for the diagnosis and prevention of some diseases. In recent years, the development of portable wearable electronic products has provided convenience for non-invasive continuous monitoring of physiological signals, and the monitoring of pulse signals is a representative one, which has been widely used in the assessment of cardiovascular health. Hypertension is the main factor inducing cardiovascular disease, which is prevalent in middle-aged and elderly people. Normal blood pressure is the premise of blood circulation and normal metabolism. Blood pressure below or above the normal level will cause serious consequences, even life-threatening. Therefore, real-time monitoring of blood pressure is an important development direction in the field of POCT.
4.1.1 Piezoelectric sensing
Piezoelectric sensor is a kind of sensor based on piezoelectric effect, which is mainly made of piezoelectric materials. When the piezoelectric material is stressed, the surface generates charges to realize the force-electricity conversion, and the strain information is evaluated by the collected electrical signals. The fingertip is one of the most common parts of the human body in contact with external objects, and the fingertip pulse waveform not only contains a wealth of physiological information, but also facilitates the direct extraction of physiological parameters. It is of great significance to enable objects such as tables, keyboards, mice and mobile phones, which are often in contact with fingertips, to sense epidermal pulse signals for accurate health assessment without time and place constraints. Meng et al. Proposed a pressure sensor based on an interlocking nanohemispherical layered structure (Fig. 10a), with polyethylene (PE) and polydimethylsiloxane (PDMS) in the form of nanohemispherical array as the tribolayer and gold layer as the electrode (Fig. 10B)
[72]. This structure enables the sensor to accurately extract the weak pulse wave signal from the fingertip artery on the surface of the object. When the fingertip is in contact with the sensor, the pulse pressure generated during the systolic phase of the pulse prompts the PE and PDMS nanohemispherical layers to begin relative motion. Because the two materials have different electron affinities, under electrostatic induction, opposite charge accumulation is generated, PE gets electrons, PDMS loses electrons, one side of PDMS generates positive friction charge, and one side of PE generates negative friction charge, so that a potential difference is generated on the two electrodes. When the fingertip pulse enters diastole, the two triboelectric layers separate and a gap appears, which causes the potential difference between the electrodes to drop, eventually reaching a state of charge equilibrium. Subsequently, as the fingertip pulse enters systole again, the PE and PDMS tribolayers come back into contact, and countercurrent flow develops. Therefore, a regular alternating current signal and a periodic frictional voltage are generated in response to the pulse pressure. The sensor has a sensitivity of 49.8 mV/Pa, a response time of less than 6 ms, and a durability of more than 4 months. Its excellent performance makes it a promising alternative to complex medical devices, based on the field of cardiovascular wireless monitoring.
With the increasing application of electronic devices in the field of sensing, the amount of electronic waste is also rising, and some toxic materials have caused serious environmental pollution. Therefore, there is an urgent need to develop degradable POCT devices that use naturally degradable sensing materials and substrates and reduce adverse effects on the environment. Guo et al. Proposed a flexible degradable pressure sensor fabricated by sandwiching porous MXene-impregnated tissue paper between interleaved electrode-modified biodegradable polylactic acid (PLA) sheets (Fig. 10 C)
[73]. Among them, MXene is a two-dimensional layered structure with large specific surface area and high conductivity, which has been widely used in electrochemical energy devices. The sensor has high sensitivity, the detection limit is as low as 10. 2 Pa, the detection range is up to 30 kPa, the response time is about 11 ms, the power consumption is only 10
-8W, and the sensor has good durability and is degradable, so it is an environmentally friendly product. This achievement can be used to predict the potential health status of patients and map tactile stimuli as electronic skin, showing potential in personal health care, clinical diagnosis and next-generation artificial skin.
4.1.2 Photoelectric sensing
Photoelectric sensing is a sensing technology based on the principle of photoelectric effect, which can convert the change of the measured signal into the change of the optical signal, and then further convert it into the electrical signal with the help of photoelectric elements. Photoplethysmography (PPG) is a non-invasive method to detect the change of blood volume in living tissue by means of photoelectric sensing technology, which can record the change of blood volume in the beating cycle by using the different intensity of reflected light after the light signal is absorbed by human blood and tissue. Chandrasekhar et al. used PPG technology to develop a smartphone-based cuffless blood pressure monitoring device
[74]. The device has a case made by 3D printing technology, which can be mounted to the back of the phone, with a PPG, pressure sensor and data processing module to measure blood volume fluctuations and applied finger pressure, and transmit the measurements to the smartphone. The app on the smartphone calculates systolic and diastolic blood pressure from the data provided by the sensors, visualizing the information back to the user. The blood pressure measurement result of the device is equivalent to that of the traditional detection device, and the device is more convenient to carry, can monitor the blood pressure anytime and anywhere, and is suitable for middle-aged and elderly people.
4.2 Body temperature
Body temperature is the product of material metabolism into heat energy. The body temperature of healthy people is relatively constant, thanks to the thermoregulatory center of the hypothalamus. Maintaining a constant body temperature is a necessary condition to ensure the normal operation of life activities. The phenomenon of body temperature exceeding the normal range is called fever, which is a clinical manifestation of many diseases and is also included in the early symptoms of COVID-19. Therefore, real-time monitoring of body temperature can help people find hidden dangers of diseases and see a doctor in time.
4.2.1 Thermopile sensing
Thermopile is a kind of pyroelectric infrared sensor, which is a device composed of thermocouples. It is widely used in ear thermometer, radiation thermometer, electric oven, food temperature detection and other fields. Core body temperature is one of the key signs to diagnose and prevent severe climate injury (such as low temperature, high fever and heatstroke), especially during the epidemic period, it is a key indicator to examine the health status of the human body, and it is essential to monitor the patient's condition. Chaglla et al. Designed a portable tympanometer (Fig. 11 A) that can be used to measure core body temperature, and solved the lack of accuracy of traditional tympanometers by coating graphene sheets on the lens of the infrared thermopile sensor
[75]. The device is based on the study of ear canal geometry, manufactured using computer-aided design software and three-dimensional printing technology, and has a Bluetooth module that can display and plot raw data on a smartphone application (Figure 11 B). This device has great potential in the field of personalized medicine.
4.2.2 Thermal imaging testing
Thermal imaging is a detection technology that converts heat into electrical signals, then generates thermal images, and calculates the temperature value. During the COVID-19 outbreak, reducing person-to-person contact can help contain the spread of the virus. Therefore, thermography is widely used because it can quickly measure the temperature distribution without physical contact. At present, there are many thermal camera devices on the market that apply thermal imaging technology to smartphones, which are low-cost and portable, and can measure body temperature anytime and anywhere, but the detection accuracy is lacking. In body temperature detection, although the absolute temperature fluctuates frequently, the relative temperature in a single thermal image is highly reliable, which is a characteristic of a thermal camera. Yoshikawa et al. took full advantage of this feature to propose the concept of a thermal bracelet, which is a method for dynamic offset correction of thermal images captured by smartphones (Figure 11C)
[76]. First, the thermal bracelet takes a thermal image of a reference point (wrist or palm) and measures its absolute temperature. Subsequently, the offset of the reference point is estimated by comparing the temperatures measured by the thermal camera and the thermal bracelet; Finally, the thermal bracelet corrects the temperature in the thermal image according to the offset. The experimental results show that the mean absolute error and standard deviation of the thermal camera measurement are reduced by 49. 4% and 64. 9%, respectively, and the Pearson correlation coefficient is increased by 112% at room temperature, which highlights the effectiveness of the thermal band. Thermal imaging technology has been widely used in various health care and advanced research fields. This achievement solves the problem of large errors in low-cost thermal cameras and may become a new development thrust.
4.3 Heart rate and respiration
Heart rate refers to the number of beats per minute. In a quiet state, the heart rate of a normal person is generally 60 to 100 beats per minute. Heart rate changes are closely related to heart disease. The heart rate of patients with heart disease is usually not within the normal range. Monitoring the heart rate can help to detect the hidden dangers of heart disease so as to treat them in time. Respiratory behavior contains the key parameters of lung function. At present, the commonly used respiratory behavior detection method is to use spirometer for spirometry. Spirometer is large and expensive. Generally, it needs to be operated with the assistance of professional medical staff, and can not be monitored continuously. Therefore, the development of respiratory monitoring-based POCT devices is helpful for the assessment of respiratory diseases.
4.3.1 Piezoelectric sensing
Because heart rate and respiratory rate can provide critical information about health status, smart mobile devices to monitor heart rate and respiration are a hot development of POCT. At present, although various traditional heart rate sensors have strong applicability, they are cumbersome to operate, have poor portability, and rely on external power supply, which brings bad experience to users and limits their application in mobile medical products. In recent years, batteryless flexible pressure sensors have opened up the market in the field of monitoring respiration and heartbeat. However, most of the traditional flexible batteryless pressure sensors need to be closely attached to the skin when detecting subtle physiological signals, which will bring uncomfortable feelings. Chen et al. Developed a set of non-contact heart rate and respiration monitoring equipment (Fig. 12 A), and adopted the self-powered pressure sensor with flexible hollow microstructure (HM-SPS) for the first time, which further improved the sensing performance (Fig. 12 B)
[77]. The sensor is actually a nanogenerator made of pressure-sensitive material with an encapsulated sheath-core structure, the outer sheath is made of folded fluorinated ethylene propylene and silver sheet, and the inner core is made of folded ethylene-vinyl acetate copolymer with hollow microstructure and silver sheet. The system has the advantages of low manufacturing cost and 18.98 V·kPa
-1 of detection accuracy, can continuously and stably monitor the heart rate and the respiratory rate in a non-contact mode, and can transmit the information to a mobile phone for a user to obtain conveniently. The research has big implications for mobile health, professional motion detection, artificial intelligence and wearable electronics.
4.3.2 Ultrasonic sensing
Chen et al. Proposed a set of wireless wearable sensing devices capable of continuously monitoring respiratory behavior, which can transmit respiratory information to a smartphone in real time (Fig. 12c)
[78]. The device consists of three subsystems: an ultrasonic transmitter, an ultrasonic receiver, and a data collector. The sensor measures the change of the distance between the transmitter and the receiver, and converts the linear strain of the wearing part into the change of the lung volume. The lung volume change data is wirelessly transmitted to the smartphone through the data collector, and calculated by the customized application program to realize the continuous real-time monitoring of respiratory behavior. The evaluation results showed that the average deviation of the sensor from the spirometer measurement was less than 4.25%. Continuous measurement of respiratory behavioral parameters helps to track chronic respiratory disease so that patients can seek appropriate treatment.
5 Conclusion and prospect
In this paper, we systematically review the application of various smartphone-based sensors in POCT in the past three years, taking different detection objects as the classification basis, combined with the corresponding sensing methods, and outline their advantages and disadvantages (Table 1). These sensing devices effectively overcome the limitations of traditional laboratory testing, especially in resource-limited areas, and have great potential. However, in order to better develop smartphone-based POCT, some key issues need to be addressed: (1) The detection efficiency of POCT should be guaranteed. POCT is a rapid detection method, which can help patients get the results in time. If the detection efficiency is not improved, POCT will lose its significance for development. In this regard, multi-channel detection technology has great development value, so that sensing equipment can process multiple samples at one time, which can greatly improve efficiency while ensuring accuracy. (2) Cost is an important factor affecting promotion. Reducing the cost of materials and reagent consumption is the key to realize the nationalization of health care. The emergence of 3D printing technology has played a certain role in reducing the consumption of materials, but it is still a relatively expensive technology, and there is a long way to go to continue to explore ways to reduce costs. (3) Mobile applications should be more practical. Because all phone-based POCT systems require specialized applications, the popularity of an application directly affects the fate of a POCT device. Especially for some sensing devices mainly aimed at middle-aged and elderly users, such as hypertension detection devices, the application should be as simple and easy to use as possible. (4) The test sample needs to be diversified. Selecting only one target analyte as a diagnostic indicator is far from enough in modern medicine, and diversified tests can help health departments to quickly and accurately formulate plans to deal with infectious disease outbreaks. Body fluid detection still occupies the mainstream position in POCT, but no matter what kind of body fluid detection, there are more or less shortcomings. Xu et al. Proposed a sensing patch suitable for sweat, blood, tears, urine and other body fluids, which can detect calcium and chloride ions in the human body. This patch has strong portability and provides a new idea for the development of body fluid detection
[79]. In recent years, POCT technology based on exhaled VOCs and vital signs has developed rapidly, which may be a suitable alternative in the case of inconvenient access to body fluids
[80]. (5) Study new energy supply schemes. POCT equipment should have the characteristics of portability, high integration and integration, and it is a consistent pursuit to get rid of heavy batteries. Self-powered sensors have the advantages of low cost, environmental protection and portability, which can replace batteries to a certain extent by converting all kinds of energy in the environment into electricity. With the improvement of nano-friction generator technology, many wearable self-powered sensing devices have been designed, and the improvement of energy supply efficiency and wearing comfort will bring new markets to POCT technology
[81⇓~83][84,85].
In a word, the opportunities and challenges faced by POCT based on smart phones are enormous. Looking back on its development process, technologies such as nanotechnology, flexible electronics, self-powered, microfluidics, multi-channel detection and biodegradability have helped researchers overcome many problems.It is the future development trend to combine machine learning, big data analysis and other emerging disciplines to make the whole system more intelligent
[86][87][88][89⇓~91][92]. The introduction of more promising technologies and early commercialization need the joint efforts of researchers.
National Natural Science Foundation of China(21890742)
National Natural Science Foundation of China(21727815)
Natural Science Foundation of Shandong Province(ZR2020QB092)
Postdoctoral Applied Research Project of Qingdao(202142)
State Key Laboratory of Bio-Fibers and Eco-Textiles(ZKT23)
State Key Laboratory of Bio-Fibers and Eco-Textiles(GZRC202025)
PDF
