Introduction

The outbreak of contagious diseases, especially those like coronavirus disease 2019 (COVID-19), presents substantial risks due to the high potential for transmission during direct contact between medical teams and patients. Healthcare workers on the front lines can face significant exposure to the virus, even with the use of personal protective equipment, owing to the possibility of asymptomatic spread. This not only endangers their health but also poses a risk of further transmission to other vulnerable patients within the healthcare setting, potentially turning medical facilities into viral hotspots. Moreover, infected healthcare professionals can inadvertently contribute to community spread, amplifying the impact of the pandemic and placing additional strain on healthcare resources and systems.

Owing to the increasing number of infected people, the medical supplies and healthcare personnel needed during the COVID-19 pandemic could not meet the surge in medical demand. These patients needed to be quarantined properly and promptly, and examinations and tests could not be performed adequately. During the COVID-19 pandemic, more effective measures to prioritize patients were needed, and most care centers relied only on simple chest X-rays, oxygen saturation monitors, thermometers, and patient-reported symptoms. If there were a means to assess lung parenchymal involvement in patients during COVID-19, such as by using an auscultation device, many of them would have been prioritized for necessary treatments and care.

Moreover, population aging and declining health over the next 2 decades will pose significant challenges. The healthcare gap resulting from an increase in the elderly population per capita is unavoidable. According to the Korean Statistical Office, the population aged 65 years and above will exceed 10 million by 2025. This is approximately 20.3% of the total population of South Korea. However, projections indicate that by 2030, this demographic will constitute one-third of the total population, and by 2070, it will increase to 47.5% [1]. The number of medical professionals remains limited, and the number of patients that each medical professional must attend to is increasing. Ensuring universal one-on-one medical consultations with physicians is exceedingly challenging.

Therefore, it is essential to develop a diagnostic method that allows healthcare professionals to efficiently and conveniently evaluate an individual’s health status in real time, regardless of location, while ensuring the safety of the medical teams. Additionally, there is a need to address the gaps in healthcare caused by the shortage of medical professionals due to the aging population.

Smartphone-based patient monitoring systems studied in the past

We investigated several previous studies on the use of smartphones in patient monitoring and found three significant smartphone-based patient monitoring systems.

In 2018, a team in Mexico presented an automated mobile health checkup system comprising an acoustic sensor, a smartphone device, and an Android operating system-based application [2]. This portable healthcare setup automatically identified crackling sounds using a sound-capturing device and smartphone application. The team’s motivation was to overcome the limitations of detecting irregular respiratory sounds such as crackles, which are often associated with respiratory ailments. The accuracy of the system for identifying fine crackles ranged from 84.86% to 89.16%, with sensitivities between 93.45% and 97.65% and specificities between 99.82% and 99.84%.

In 2021, Chinese researchers developed Auscal Pi, a low-cost, ear-contactless electronic stethoscope [3]. This system operates on Raspberry Pi and integrates a chest piece for auscultation, battery, display, and speakers onto a single printed circuit board. Medical personnel can use the chest piece on various parts of the patient’s chest, as usual. The sounds can be played back in real-time using a micro speaker or saved for subsequent analysis. Compared to the U.S. Food and Drug Administration (FDA)-approved 3M Littmann Stethoscope (3M, St. Paul, MN, USA), Auscal Pi showed a significant correlation coefficient ranging from 0.3449 to 0.4797 (p<0.001). Notably, it utilizes a standard USB-attachable microphone.

In 2023, American researchers proposed an artificial intelligence (AI)-assisted auscultation system called StethAid (Auscultech Dx LLC), designed for iPhone operating system (iOS) mobile devices [4]. The system consists of a wireless electronic stethoscope, an iOS smartphone, and an application. Once the data are collected, they are stored on the patient’s phone and, upon connection to Wi-Fi, are transferred to the cloud for analysis using AI algorithms. The research team employed two deep learning algorithms, ResNet18 and Harmonic Networks, for the analysis. The results showed higher accuracy with Harmonic Networks in terms of sensitivity (77% vs. 78%), specificity (70% vs. 86%), and accuracy (74% vs. 84%).

Many other studies have used smartphones to monitor health conditions or assess the heart or lungs. However, they all require additional components, such as chest pieces, to enhance the sound quality, and smartphones merely capture and sort data for later analysis by medical teams. Additional modules, whether few or many, incur additional costs that can hinder the adoption of smartphones as digital stethoscopes. With the integration of on-device AI chips, smartphones can potentially function as smart stethoscopes. This implies that smartphones can independently analyze health data and provide immediate health reports using on-device AI chips.

Viability of smartphone-based digital stethoscopes with artificial intelligence

1. Analysis of auscultation accuracy using built-in smartphone microphones for audible wavelengths

Researchers in Portugal conducted a study comparing lung auscultations performed using a stethoscope with those conducted using a smartphone microphone [5]. This observational cross-sectional study included 134 patients from the pediatric and pulmonology departments of a tertiary care facility. The cohort included patients with cystic fibrosis (31%), other respiratory diseases (29%), asthma (28%), and no respiratory diseases (12%), with a median age of 16 years and male predominance of 54%. The participants underwent traditional auscultation at four predetermined chest locations, followed by smartphone auscultation at the same site. The smartphone recordings were independently analyzed for the quality and presence of adventitious sounds, including wheezes and crackles.

Of the 1,060 smartphone recordings, 769 (73%) were deemed of sufficient quality for analysis. Quality assessment revealed a higher proportion of acceptable recordings from the trachea (82%) than from other auscultation sites. Adventitious sounds were detected in 35% of the participants, representing 14% of the quality recordings. The agreement between the conventional and smartphone auscultation methodologies, as measured by Cohen’s kappa coefficient, was 0.35, indicating a fair level of concordance in detecting adventitious sounds.

These findings suggest that smartphone-based lung auscultation is a feasible approach in clinical settings, offering significant agreement with traditional stethoscope assessments. The high proportion of quality recordings underscores the potential of smartphone microphones in capturing clinically relevant lung sounds. However, the fair level of agreement on adventitious sound detection highlights the need for further refinement and standardization of smartphone auscultation techniques.

2. Auscultation of inaudible wavelengths from the body, an old frontier made possible with digital devices

It was inevitable that auscultation techniques would develop throughout the history of medicine because physicians depended on their ears for auscultation [6]. In addition, non-audible wavelengths from the body, mainly arising from blood vessels and gut movements, were discovered and researched as early as the 1950s without making them relevant to clinical implications. Some of these discoveries opened possibilities in medical fields under the term “ultrasound” [7]. Currently, ultrasound techniques are used for the joints, heart, and almost every part of the human body [8], including the brain [9]. The Focused Ultrasound Foundation developed a technique that applies specific wavelengths starting at multiple spots, which converge at a targeted point inside the human skull to generate sufficient energy to destroy a tumor or stimulate various functions at the brain site. An early version of the ultrasound technique converted ultrasound to an audible sound such as the heart sound from a fetus. It then evolved into M-mode, which is a graphic display of a linear signal with a time axis. M-mode is still used to measure valvular functions of the heart [10]. This technique has evolved to provide dynamic reconstructed three-dimensional (3D) images of fetuses without complications.

If a technique to decipher the characteristic wave was available, infrasound could also help detect subtle changes, even at the cellular level [11]. When sufficient digital data are gathered with the help of AI, it would be possible to isolate relevant waves from infrasound that are currently considered as background noise. The authors expect to contribute to the development of the infrasound technique from audible sound to dynamic 3D videos, just as ultrasound machines generate 3D live-streaming images [12-14].

3. Several examples of artificial intelligence advancement in medical sectors

AI is opening new possibilities in medical diagnostics by enhancing the accuracy and functionality of smartphone-based digital stethoscopes [4,15-20]. Smartphone-based digital stethoscopes address the limitations of traditional stethoscopes, which rely heavily on user expertise and are often affected by environmental factors [19].

In 2019, Chowdhury et al. [21] explored the potential of a smartphone-based digital stethoscope system equipped with AI to analyze heart sounds and detect abnormalities. Their work utilized over 10,000 recordings from the PhysioNet Challenge 2016 dataset and extracted 27 key features from the time, frequency, and mel-frequency cepstral coefficient domains. Among the 22 machine learning models tested, the ensemble classifiers performed the best, achieving an accuracy of 97% for normal heart sounds and 88% for abnormal sounds. While the system demonstrated the potential to improve early cardiovascular diagnostics, challenges included reliance on personal computer-based processing, an imbalance in training data, and an inability to guarantee absolute diagnostic accuracy, which raises questions about clinical trust.

In 2022, Ghanayim et al. [22] further explored the role of AI in smartphone-based digital stethoscopes with the development of VoqX (Sanolla), designed to diagnose aortic stenosis (AS). By capturing heart sounds in the 3 to 2,000 Hz range, the AI analyzed features, such as ejection time and signal entropy, to identify moderate or severe AS. In testing, the AI demonstrated sensitivities of 86% and 90% and specificities of 100% and 84% in training/validation and independent testing, respectively. The sensitivity increased to 93% for severe AS but remained lower for mild AS (55%), highlighting the potential for missed diagnoses. External factors, such as noisy environments and patient conditions, such as obesity, posed further challenges.

AI-enhanced digital stethoscopes address the limitations of traditional auscultation by digitizing and analyzing cardiopulmonary sounds in real time, offering consistent and accurate diagnoses [22] for conditions such as abnormal breath sounds [23] or heart sounds, while supporting remote monitoring and chronic care through cloud-based systems [20,24].

To support the adoption of AI in medical devices such as smartphone-based stethoscopes, the FDA has worked to reduce the regulatory burden, particularly for assistive tools [25]. The FDA classifies computer-assisted detection devices, designed to highlight features for clinician review rather than autonomous diagnosis, as lower-risk tools that require fewer regulatory hurdles. This approach encourages innovation, while ensuring safety. For instance, the FDA has approved AI applications in radiology such as computed tomography (CT)-based stroke notification systems and tools for identifying pneumothorax or cancerous lesions in imaging. These examples demonstrate the ability of AI to enhance the diagnostic workflow without replacing human oversight.

Although these developments are significant advancements, the inability to guarantee 100% diagnostic accuracy introduces risks. Missed or incorrect diagnoses can delay or compromise care, raising concerns about patient safety and liability. These challenges are compounded by legal and insurance issues, as it remains unclear who bears the responsibility for AI-related errors. Clear frameworks for risk sharing and accountability are critical for fostering clinical confidence and supporting adoption.

The integration of AI into smartphone-based digital stethoscopes and other assistive medical tools has immense potential for improving diagnostic precision and accessibility. However, addressing technical limitations, building trust in the technology, and developing robust legal [26] and insurance frameworks, combined with regulatory support from the FDA, will be essential to fully realize their capabilities in clinical practice.

Hurdles for applying advancing technology to current healthcare

1. Regulations

Implementation of new technologies in the healthcare industry has created a complex set of regulatory challenges. As AI and other advanced solutions are integrated into healthcare systems, issues related to data privacy, security, and information sharing have become increasingly prevalent [27,28]. Policymakers and regulatory bodies are tasked with developing a comprehensive framework that can effectively address the multifaceted nature of these challenges while also ensuring that patient’s rights and the integrity of healthcare data are protected.

One of the primary concerns is the protection of individuals’ medical records and other personal health information as outlined in the Health Insurance Portability and Accountability Act Privacy Rule of the United States [28]. However, as the healthcare industry continues to evolve, existing regulations have struggled to keep pace with rapid technological advancements, posing challenges for the enforcement and development of technical standards [28].

Indeed, the transition from paper-based to electronic health records has been a major focus of federal regulations in the healthcare sector for nearly 2 decades. The introduction of mandatory insurance, federal and state health insurance exchanges, and expanded Medicaid programs under the Patient Protection and Affordable Care Act have further increased regulatory oversight in the healthcare industry [29].

2. Limitations in generating big data and data storage capacity, along with computing power, in artificial intelligence technology

In an environment where sensors continuously generate large volumes of digital data, AI plays an important role in storing, analyzing, and interpreting these datasets [30,31]. Although AI facilitates meaningful data combinations, there are inherent limitations in analyzing its algorithms and assessing causality, reliability, reproducibility, and the potential influence of biases during decision-making. These challenges [32,33] necessitate a careful and deliberate approach, particularly when applying AI to human-centered applications.

Proposals for improving the data storage infrastructure to handle such demands have been made [34,35]; however, challenges persist in efficiently storing and retrieving the real-time data generated by sensors. For example, in the case of audio data, sensors record continuous data streams for each wavelength on a millisecond timescale. When inaudible wavelengths are included, the amount of data increases. According to the Nyquist–Shannon sampling theorem, reconstructing a signal accurately requires a sampling frequency that is at least twice the maximum signal frequency [36]. This inevitably leads to an increase in data volume [37].

In addition, the real-time duplication of data channels from multiple auscultation sites contributes to the data size. For example, routine chest auscultation typically includes 32 anterior and posterior sites, heart auscultation involves four sites, and abdominal auscultation includes two sites. If new auscultation modalities such as egophony are integrated or additional sites are examined, the data volume will further increase. Moreover, distinguishing and measuring environmental noise alongside human body sounds is essential, adding to the overall data size. These factors highlight the need for efficient systems to manage the rapidly growing volume of data generated by such applications.

3. Proposed approach to overcoming such hurdles: implementation in veterinary medicine first

In veterinary medicine, the early detection of diseases in animals is challenging, as animals cannot verbally express the location of discomfort or describe their symptoms. Consequently, regular veterinary visits for physical examinations are essential for monitoring and maintaining animal health. The use of smart devices for patient monitoring, particularly smartphone-based technologies, is a growing area of veterinary medicine research. For instance, smartphone-based electrocardiography (ECG) has demonstrated effectiveness comparable to that of the standard ECG in various species, including dogs [38], sheep [39,40], donkeys [41], cows [42], and horses [43-45].

In 2023, Vezzosi et al. [45] conducted a study with 99 dogs and nine cats to assess the accuracy and feasibility of a smartphone-based stethoscope from Eko Devices, Inc. Their findings indicated that the digital stethoscope effectively detected abnormalities, such as heart murmurs, gallop sounds, ventricular premature complexes, and bundle branch blocks, without false negatives. However, insufficient data was collected from 10% of the animals owing to noncompliance.

The use of smartphone-based digital stethoscopes by owners has significant potential in veterinary medicine. In companion animals, this technology can improve the accuracy of auscultation by enhancing their compliance and reducing stress, as animals are typically more relaxed when examined by their owners at home [46-48]. At-home monitoring could also help ensure timely medical intervention, allowing owners to seek veterinary care promptly when needed.

In the context of industrial animal management, smartphone-based digital stethoscopes could be beneficial for controlling the spread of infectious diseases. By enabling early detection of abnormal sounds in individual animals, this technology could aid in the prompt identification of potentially infected animals, thereby contributing to herd management [49,50].

However, several limitations should be addressed when applying these technologies in veterinary medicine: the control of animal behavior during recordings, anatomical differences across species, and the need for species-specific datasets to train AI models. In addition, ensuring the quality and standardization of the data collected by animal owners or farm workers is a significant hurdle. Addressing these challenges is essential for the successful integration of smartphone-based digital stethoscopes into routine veterinary practice.

Conclusion

By 2029, it is projected that 6.4 million people will own smartphones (Fig. 1) [51,52]. By leveraging this widespread accessibility, the establishment of a healthcare system that utilizes smartphones for initial and regular diagnoses, traditionally conducted using stethoscopes, CT scans, and other methods, can significantly enhance healthcare efficiency. Smartphone-based digital stethoscopes enhanced with AI present an opportunity to address the current gaps in healthcare diagnostics, especially during pandemics and in resource-limited environments. These devices, which leverage smartphone microphones and integrated AI, offer real-time auscultation without relying on additional hardware. However, challenges such as regulatory compliance, diagnostic accuracy, and data management remain barriers to widespread implementation. Refining these technologies through pilot applications in veterinary medicine and enhancing data standardization could help overcome these issues. Although not without limitations, smartphone-based AI stethoscopes have potential as practical tools for improving accessibility and efficiency in diagnostic practices.

Article information

Conflicts of interest

No potential conflict of interest relevant to this article was reported.

Funding

None.

Author contributions

Conceptualization, Data curation, Formal analysis, Resources: HL, JSB; Methodology, Visualization, Validation: HL; Investigation: HL, GK, Project administration, Supervision: JSB; Writing-original draft: HL, GK; Writing-review & editing: all authors.

Fig. 1.

Worldwide population by age group and projected smartphone users (2017–2028).

References

• 1. Statistics Korea. Future population projections. Key population indicators (sex ratio, population growth rate, population structure, dependency ratio, etc.)/nationwide [Internet]. Korean Statistical Information Service; 2024 [cited 2024 May 23]. https://kosis.kr/statHtml/statHtml.do?orgId=101&tblId=DT_1BPA002&vw_cd=MT_ZTITLE&list_id=A41_10&scrId=&seqNo=&lang_mode=ko&obj_var_id=&itm_id=&conn_path=MT_ZTITLE&path=%252FstatisticsList%252FstatisticsListIndex.do.
• 2. Reyes BA, Olvera-Montes N, Charleston-Villalobos S, González-Camarena R, Mejía-Ávila M, Aljama-Corrales T. A smartphone-based system for automated bedside detection of crackle sounds in diffuse interstitial pneumonia patients. Sensors (Basel) 2018;18:3813.ArticlePubMedPMC
• 3. Yang C, Zhang W, Pang Z, Zhang J, Zou D, Zhang X, et al. A low-cost, ear-contactless electronic stethoscope powered by Raspberry Pi for auscultation of patients with COVID-19: prototype development and feasibility study. JMIR Med Inform 2021;9:e22753. ArticlePubMedPMC
• 4. Arjoune Y, Salvador T, Nguyen TN, Telluri A, John T, Schroder JC, et al. STETHAID: an electronic stethoscope connected to iOS mobile apps for AI-assisted auscultation. In: Proceedings of the 2023 Design of Medical Devices Conference. Minneapolis, MN; April 17-19, 2023. New York: American Society of Mechanical Engineers; 2023. Paper No: DMD2023-8881, V001T05A003.ArticlePDF
• 5. Ferreira-Cardoso H, Jácome C, Silva S, Amorim A, Redondo MT, Fontoura-Matias J, et al. Lung auscultation using the smartphone-feasibility study in real-world clinical practice. Sensors (Basel) 2021;21:4931.ArticlePubMedPMC
• 6. Murphy RL. In defense of the stethoscope. Respir Care 2008;53:355–69.PubMed
• 7. Levi S. The history of ultrasound in gynecology 1950-1980. Ultrasound Med Biol 1997;23:481–552.ArticlePubMed
• 8. Tagliafico A, Bignotti B, Rossi F, Sconfienza LM, Messina C, Martinoli C. Ultrasound of the hip joint, soft tissues, and nerves. Semin Musculoskelet Radiol 2017;21:582–8.ArticlePubMed
• 9. Qiu W, Bouakaz A, Konofagou EE, Zheng H. Ultrasound for the brain: a review of physical and engineering principles, and clinical applications. IEEE Trans Ultrason Ferroelectr Freq Control 2021;68:6–20.ArticlePubMed
• 10. Feigenbaum H. Role of M-mode technique in today's echocardiography. J Am Soc Echocardiogr 2010;23:240-57; 335-7. ArticlePubMed
• 11. Chaban R, Ghazy A, Georgiade E, Stumpf N, Vahl CF. Negative effect of high-level infrasound on human myocardial contractility: in-vitro controlled experiment. Noise Health 2021;23:57–66.PubMedPMC
• 12. Hultberg B, Lundblad A, Masson PK, Ockerman PA. Specificity studies on alpha-mannosidases using oligosaccharides from mannosidosis urine as substrates. Biochim Biophys Acta 1975;410:156–63.PubMed
• 13. Makar AB, McMartin KE, Palese M, Tephly TR. Formate assay in body fluids: application in methanol poisoning. Biochem Med 1975;13:117–26.ArticlePubMed
• 14. Liteplo AS, Noble VE, Attwood B. Real-time video transmission of ultrasound images to an iPhone. Crit Ultrasound J 2010;1:105–10.ArticlePDF
• 15. Zander D, Hüske S, Hoffmann B, Cui XW, Dong Y, Lim A, et al. Ultrasound image optimization (“Knobology”): B-mode. Ultrasound Int Open 2020;6:E14–24.ArticlePubMedPMC
• 16. Sinharay A, Ghosh D, Deshpande P, Alam S, Banerjee R, Pal A. Smartphone-based digital stethoscope for connected health: a direct acoustic coupling technique. In: Proceedings of the 2016 IEEE First International Conference on Connected Health: Applications, Systems and Engineering Technologies (CHASE). Washington, DC; June 27-29, 2016. Piscataway, NJ: IEEE; 2016. p. 193-8.Article
• 17. Zhang M, Li M, Guo L, Liu J. A low-cost AI-empowered stethoscope and a lightweight model for detecting cardiac and respiratory diseases from lung and heart auscultation sounds. Sensors (Basel) 2023;23:2591.ArticlePubMedPMC
• 18. Sharma S, Pandey S, Shah D. Enhancing medical diagnosis with AI: a focus on respiratory disease detection. Indian J Community Med 2023;48:709–14.ArticlePubMedPMC
• 19. Arjoune Y, Nguyen TN, Doroshow RW, Shekhar R. Technical characterisation of digital stethoscopes: towards scalable artificial intelligence-based auscultation. J Med Eng Technol 2023;47:165–78.ArticlePubMedPMC
• 20. Chang KC, Huang JW, Wu YF. Design of e-health system for heart rate and lung sound monitoring with AI-based analysis. In: Proceedings of the 2021 IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW). Penghu, Taiwan; September 15-17, 2021. IEEE; 2021. p. 1-2.Article
• 21. Chowdhury ME, Khandakar A, Alzoubi K, Mansoor S, M Tahir A, Reaz MB, et al. Real-time smart-digital stethoscope system for heart diseases monitoring. Sensors (Basel) 2019;19:2781.ArticlePubMedPMC
• 22. Ghanayim T, Lupu L, Naveh S, Bachner-Hinenzon N, Adler D, Adawi S, et al. Artificial intelligence-based stethoscope for the diagnosis of aortic stenosis. Am J Med 2022;135:1124–33.ArticlePubMed
• 23. Joshi RC, Khan JS, Pathak VK, Dutta MK. AI-CardioCare: artificial intelligence-based device for cardiac health monitoring. IEEE Trans Hum-Mach Syst 2022;52:1292–302.Article
• 24. Wu YC, Han CC, Chang CS, Chang FL, Chen SF, Shieh TY, et al. Development of an electronic stethoscope and a classification algorithm for cardiopulmonary sounds. Sensors (Basel) 2022;22:4263.ArticlePubMedPMC
• 25. Oliynik V. On potential effectiveness of integration of 3M Littmann 3200 electronic stethoscopes into third-party diagnostic systems with auscultation signal processing. In: Proceedings of the 2015 IEEE 35th International Conference on Electronics and Nanotechnology (ELNANO). Kyiv, Ukraine; April 21-24, 2015. IEEE; 2015. p. 417-21.Article
• 26. Harvey HB, Gowda V. How the FDA regulates AI. Acad Radiol 2020;27:58–61.ArticlePubMed
• 27. Ganapathy K. Artificial intelligence and healthcare: regulatory and legal concerns. Telehealth Med Today 2021;6:252.Article
• 28. Hlávka JP. Security, privacy, and information-sharing aspects of healthcare artificial intelligence. In: Bohr A, Memarzadeh K, editors. Artificial intelligence in healthcare. Amsterdam: Academic Press; 2020. p. 235–70.Article
• 29. Kim RH, Gaukler GM, Lee CW. Improving healthcare quality: a technological and managerial innovation perspective. Technol Forecast Soc Change 2016;113(pt B):373–8.Article
• 30. Anthony DL, Appari A, Johnson ME. Institutionalizing HIPAA compliance: organizations and competing logics in U.S. health care. J Health Soc Behav 2014;55:108–24.ArticlePubMedPDF
• 31. Manjulatha B, Pabboju S. Big data analytics and its applications. Int J Recent Technol Eng 2019;8:10928–31.Article
• 32. Sivarajah U, Kamal MM, Irani Z, Weerakkody V. Critical analysis of big data challenges and analytical methods. J Bus Res 2017;70:263–86.Article
• 33. Srinivasan R, Chander A. Biases in AI systems: a survey for practitioners. Queue 2021;19:45–64.Article
• 34. Nishant R, Schneckenberg D, Ravishankar M. The formal rationality of artificial intelligence-based algorithms and the problem of bias. J Inf Technol 2024;39:19–40.ArticlePDF
• 35. Siddiqa A, Karim A, Gani A. Big data storage technologies: a survey. Front Inf Technol Electron Eng 2017;18:1040–70.ArticlePDF
• 36. Strohbach M, Daubert J, Ravkin H, Lischka M. Big data storage. In: Cavanillas JM, Curry E, Wahlster W, editors. New horizons for a data-driven economy. Cham: Springer International Publishing; 2016. p. 119–41.
• 37. Shannon CE. Communication in the presence of noise. Proc IRE 1949;37:10–21.Article
• 38. Silber EA, Boslough M, Hocking WK, Gritsevich M, Whitaker RW. Physics of meteor-generated shock waves in the Earth’s atmosphere: a review. Adv Space Res 2018;62:489–532.Article
• 39. Vezzosi T, Buralli C, Marchesotti F, Porporato F, Tognetti R, Zini E, et al. Diagnostic accuracy of a smartphone electrocardiograph in dogs: comparison with standard 6-lead electrocardiography. Vet J 2016;216:33–7.ArticlePubMed
• 40. King A, Rolph KE, Dzikiti L, Cavanaugh SM. Overall good agreement of smartphone-based and standard base-apex electrocardiography in healthy sheep. J Am Vet Med Assoc 2023;261:1–5.ArticlePubMedPMC
• 41. Cicogna M, Rishniw M, Gobbi A, Calgaro V, Porciello F, Caivano D. Evaluation of a new smartphone-based stethoscope with one-lead electrocardiogram in healthy sheep [Preprint]. SSRN; 2024 Aug 27 [cited 2024 Nov 30]. https://doi.org/10.2139/ssrn.4938203.Article
• 42. Bozzola C, Ortolina A, Guffanti I, Alberti E, Bronzo V, Zucca E. Evaluation of a digital stethoscope for electrocardiographic recording in donkeys: preliminary results. J Equine Vet Sci 2024;135:105048.ArticlePubMed
• 43. Bonelli F, Vezzosi T, Meylan M, Nocera I, Ferrulli V, Buralli C, et al. Comparison of smartphone-based and standard base-apex electrocardiography in healthy dairy cows. J Vet Intern Med 2019;33:981–6.ArticlePubMedPMCPDF
• 44. Alberti E, Stucchi L, Pesce V, Stancari G, Ferro E, Ferrucci F, et al. Evaluation of a smartphone-based electrocardiogram device accuracy in field and in hospital conditions in horses. Vet Rec Open 2020;7:e000441. ArticlePubMedPMCPDF
• 45. Vezzosi T, Alibrandi L, Grosso G, Tognetti R. Evaluation of a new smartphone-based digital stethoscope featuring phonocardiography and electrocardiography in dogs and cats. Vet J 2023;295:105987.ArticlePubMed
• 46. Kraus MS, Rishniw M, Divers TJ, Reef VB, Gelzer AR. Utility and accuracy of a smartphone-based electrocardiogram device as compared to a standard base-apex electrocardiogram in the horse. Res Vet Sci 2019;125:141–7.ArticlePubMed
• 47. Csoltova E, Martineau M, Boissy A, Gilbert C. Behavioral and physiological reactions in dogs to a veterinary examination: owner-dog interactions improve canine well-being. Physiol Behav 2017;177:270–81.ArticlePubMed
• 48. Stellato AC, Dewey CE, Widowski TM, Niel L. Evaluation of associations between owner presence and indicators of fear in dogs during routine veterinary examinations. J Am Vet Med Assoc 2020;257:1031–40.ArticlePubMed
• 49. Bishop GT, Evans BA, Kyle KL, Kogan LR. Owner satisfaction with use of videoconferencing for recheck examinations following routine surgical sterilization in dogs. J Am Vet Med Assoc 2018;253:1151–7.ArticlePubMed
• 50. Kamel MS, Davidson JL, Verma MS. Strategies for Bovine Respiratory Disease (BRD) diagnosis and prognosis: a comprehensive overview. Animals (Basel) 2024;14:627.ArticlePubMedPMC
• 51. Statista. Number of smartphone users worldwide 2016 to 2023, with forecasts from 2023 to 2028 [Internet]. statista; 2024 [cited 2024 Nov 28]. https://www.statista.com/statistics/330695/number-of-smartphone-users-worldwide.
• 52. Our World in Data. Population by age group with projections [Internet]. Our World in Data; 2024 [cited 2024 Nov 28]. https://ourworldindata.org/grapher/population-by-age-group-with-projections.

Figure & Data

References

Citations

Citations to this article as recorded by  

• LMS-ViT: a multi-scale vision transformer approach for real-time smartphone-based skin cancer detection
  A. Anny Leema, P. Balakrishnan, G. Gopichand, G. Rajarajan
  Frontiers in Artificial Intelligence.2025;[Epub]     CrossRef