Skip to article
AHFE AHFE Discoveries AHFE International Open Access Journal
AHFE OPEN ACCESS OPEN ACCESS CONFERENCE PROCEEDINGS PAPER

Drawing Connections: Artificial Intelligence to Address Complex Health Challenges

1 Department of Pediatrics, College of Medicine, University of Saskatchewan, 103 Hospital Drive, Saskatoon, S7N 0W8, Canada ↗
2 Department of Pathology and Laboratory Medicine, College of Medicine, University of Saskatchewan, Saskatoon, Canada ↗
Publication year
Digital Object Identifier10.54941/ahfe1002111

Abstract

Pattern recognition is a cornerstone of clinical care and public health practice. Historically, advances in medicine have relied on the ability of humans to detect patterns and make inferences. Modern healthcare challenges involve vast amounts of data and a level of complexity that require additional support to understand. The advancement of Artificial Intelligence has expanded our capability to detect, understand, and address patterns that were previously beyond our grasp. Artificial Intelligence has the capability to analyze otherwise insurmountable quantities of data in order to bring meaning and clarity to patterns that were previously deemed random or unintelligible. We, therefore, aim to charter a strategic path forward for innovative applications of Artificial Intelligence technology to understand and address pressing complex health challenges. The modelling capabilities of Artificial Intelligence have allowed for the simulation of potential viral mutations, as well as the development of therapeutic agents. The predictive analyses provided by Artificial Intelligence allow for a more holistic yet precise understanding of the aging process and the progression of disease, thereby allowing the extent and timing of treatments to be optimized. It has brought a new lens through which to identify malignant cells on imaging and to decode parts of the human genome previously labelled as sequences of unknown significance. On a global scale, Artificial Intelligence has given us the opportunity to better understand and anticipate the effects of climate change on health including the effects on displacement and the potential spillover and spread of new zoonotic infection diseases. We suggest how Artificial Intelligence is beginning to re-conceptualize our understanding of health and disease. The implementation of Artificial Intelligence is a pivotal time in developmental of other modern era of medical practice and public health strategies. Appropriate utilization of these new tools requires innovative thinking, critical appraisal, and tactful resource allocation to ensure issues are addressed in a timely and feasible manner.

INTRODUCTION

Pattern recognition is a cornerstone of clinical care and public health practice. This recognition of the patterns or stigmata of disease underlies historical and modern diagnostic decision-making processes (Seitzinger et al. 2021). In the context of population health, pattern recognition involves finding epidemio- logical links among cases and risk factors of disease (Kalra et al. 2019). By seeing beyond individual clinical data points and drawing connections on a population level, clinicians and public health practitioners have made pivotal discoveries to mitigate illness and advance knowledge in the field (Seitzinger et al. 2021). However, to date, such ‘hunches’ occur on a haphazard basis and are rarely followed through on. Modern healthcare challenges involve vast amounts of data and a level of complexity that requires additional support to understand.

© 2022. Published by AHFE Open Access. All rights reserved. 151

152 Seitzinger and Kalra

Artificial Intelligence (AI) is beginning to re-conceptualize our understan- ding of health and disease. The implementation of AI is pivotal in developing the modern era of medical practice and public health strategies. In an era of rapidly evolving technological advances, definitions of pattern recogni- tion have expanded to include processes conducted by computer algorithms (Kalra et al. 2021). In the context of computer science, pattern recognition refers to the identification and classification of meaningful patterns of data input by computers based on the extraction and comparison of the chara- cteristic properties or features of the data (Seitzinger et al. 2021). AI can be described as the process in which computers simulate processes and characte- ristics of humans to accomplish a task without being specifically programmed to do so. With the ever-increasing need for more accurate and precise heal- thcare delivery, the integration of AI presents a desirable solution for various data management and interpretation challenges. The integration of AI into the modern world has rapidly gained popularity, with much of the focus and funding being diverted towards AI in healthcare (Buch et al. 2018). Every day the potential benefits of AI are being realized in a growing number of medical fields and are playing an increasing role in tackling healthcare issues. The advancement of AI has the potential to expand our capability to detect, understand, and address patterns that were previously beyond our grasp.

CLINICAL APPLICATIONS OF ARTIFICIAL INTELLIGENCE The capabilities of AI have shown promise in meeting the demands of any aspect of clinical care. Currently, AI systems are being introduced into clini- cal care at an unprecedented pace, with new applications arriving every day from health systems worldwide. These capabilities have been used to enha- nce the efficiency of documentation and detect disorders in various medical disciplines. AI systems have proven helpful in speech recognition with a voca- bulary exceeding 300,00 words with an accuracy of 99%% (Schuman 2014). Given that the radiology field depends heavily on the visual identification of mass numbers of scans, AI systems can increase workflow without com- promising the accurate identification of abnormalities (Paiva and Prevedello 2017). Using the recording of auscultation collected by electronic stetho- scopes and ambulatory electrocardiograms, AI systems have been shown to perform at a level of board-certified cardiologists in detecting murmurs and arrhythmias (Hannun et al. 2019). In radiology, AI interpretations systems have been shown to expedite the treatment of intracranial bleeding, multiple sclerosis, traumatic brain injury, pulmonary embolisms, and wrist fractures. In Pediatrics, such computer-based interpretation systems have outperformed radiologists in performing bone age assessments with increased accuracy and decreased variability (Tajmir et al. 2019) and for ophthalmologists to detect prematurity of retinopathy (Wang et al. 2018). In genetics, AI systems have shown promise in detecting over 200 syndromes by assessing facial phe- notypes through digital images with an accuracy of over 90% (Gurovich et al. 2019). These examples illustrate how the growing capabilities of AI are currently being used in individual medical disciplines.

Drawing Connections: Artificial Intelligence to Address Complex Health Challenges 153

Table 1. Accuracy of artificial intelligence in different medical disciplines.

Table available in the original PDF paper.

Discipline Capability Data Accuracy Auscultation 88% accurate (Hannun et al. 2019) Cardiology Identifies common murmurs

Ultrasound 98% accurate (Tajmir et al. 2019)

Radiology Detects Developmental Dysplasia of the Hip Dermatology Identifies skin lesions Imaging of lesion

Imaging of retina

Sensitivity: 88.9% Specificity: 75.7% (Haenssle et al. 2018) Ophthalmology Detects Retinopathy of Prematurity

Sensitivity: 84.9% Specificity: 96.9% (Wang et al. 2018) Genetics Identifies over 200 syndromes Imaging of face

91% accurate (Gurovich et al. 2019) Clinical Decision Making Diagnosing

Patient History

79-90% accurate (Liang et al. 2019)

common pediatric presentations

The ability to serve as a copilot in clinical decision-making has long been a hope and need for helathcare providers interested in AI technology. For example, AI clinical decision support systems have been shown to dia- gnose common presentations of asthma, gastroenteritis, pneumonia, sinusitis, upper respiratory infections, encephalitis and psychiatric disorders with accu- racies ranging from 70% to 90% (Liang et al. 2019). To provide context for these levels of accuracy, these systems currently outperform junior staff physicians (who have completed undergraduate studies, medical school and residency training) in clinical decision-making but remain inferior to the accu- racy of more experienced staff physicians. Recent studies on the accuracy of AI systems in different medical disciplines are summarized in Table 1.

THE NEXT FRONTIER OF ARTIFICIAL INTELLIGENCE IN MEDICINE AI has the capability to analyze otherwise insurmountable quantities of data to bring meaning and clarity to patterns that were previously deemed random or unintelligible. The modelling capabilities of AI have allowed for the simu- lation of potential viral mutations and the development of therapeutic agents. The predictive analyses provided by AIallow for a more holistic yet precise understanding of the ageing process and the progression of disease, thereby allowing the extent and timing of treatments to be optimized. Through AI, clinicians will be able to efficiently and accurately identify predictive clinical markers for educational and diagnostic purposes (Colling et al. 2019). The predictive capabilities of AI are showing promise in predicting conditions such as sepsis earlier, enabling clinicians to act more proactively rather than reactively (Goh et al. 2021). AI has the potential to inform the prevention,

154 Seitzinger and Kalra

treatment or management of existing or impending neurodegenerative dise- ases (Deng et al. 2020). It has brought a new lens to identify malignant cells and decode parts of the human genome previously labelled as sequences of unknown significance (Hie et al. 2021). Advancements in understanding the implications of differences in the human genome allow for treatments tailo- red to each patient, with a more comprehensive understanding of interactions among inherent susceptibilities, circumstances and proposed interventions. On a global scale, AI systems can be utilized to model trends and identify cor- relations (Cabitza and Banfi 2018). AI has given us the opportunity to better understand and anticipate the effects of climate change on health and predict the impacts of potential interventions (Cowls et al. 2021). AI systems have shown the potential to anticipate potential spillover events of zoonotic infe- ctious diseases and guide disease control strategies (Plowright and Hudson 2021).

STRATEGIZING A PATH FORWARD Navigating the transition to an AIassisted era of healthcare delivery will require an appreciation of the opportunities and limits of each technology. These tools have the potential to have transformative effects on screening, early diagnosis, and determining disease susceptibility and progression. For the first time in the development of modern medicine, we have the ability to recognize, understand, and address patterns that were previously deemed random or happenstance. AI alone cannot provide adequate care, and at the same time, lack of AI utilization will hinder the potential benefits that can be delivered to patients. As the demands for AI in healthcare increase, so must the capabilities of this technology. Advancing the AI system requires incor- porating more than just numerical data when assessing patient outcomes or giving a diagnosis. Appropriate utilization of these new tools requires inno- vative thinking, critical appraisal, and tactful resource allocation to ensure issues are addressed in a timely and feasible manner.

CONCLUSION

Historically, advances in medicine have relied on the ability of humans to detect patterns and make inferences. Advancements in the field of AI have expanded our capabilities to recognize patterns and draw inferences. The modelling capabilities of AI have allowed for the simulation of potential viral mutations and the development of therapeutic agents. The predictive analyses provided by AI allow for a more holistic yet precise understanding of the ageing process and the progression of disease, thereby allowing the extent and timing of treatments to be optimized. It has brought a new lens through which to identify malignant cells on imaging and to decode parts of the human genome previously labelled as sequences of unknown significa- nce. On a global scale, AI has given us the opportunity to better understand and anticipate the effects of climate change on health, including potential spillover and spread of zoonotic infectious diseases. Novel AI tools such

Drawing Connections: Artificial Intelligence to Address Complex Health Challenges 155

as these require further exploration to identify the wide range of potential applications and challenges.

REFERENCES

Buch, V.H., Ahmed, I., and Maruthappu, M. (2018), “Artificial intelligence in medicine: Current trends and future possibilities.” British Journal of General Practice [online], 68(668), 143–144. Cabitza, F., and Banfi, G. (2018), “Machine learning in laboratory medicine: Wai-

ting for the flood?.” Clinical Chemistry and Laboratory Medicine [online], 56(4), 516–524. Colling, R., Pitman, H., Oien, K., Rajpoot, N., Macklin, P., Snead, D., Sackville, T., google and Verrill, C. (2019), “Artificial intelligence in digital pathology: a roa- dmap to routine use in clinical practice.” Journal of Pathology [online], 249(2), 143–150. Cowls, J., Tsamados, A., Taddeo, M., and Floridi, L. (2021), “The AI Gambit

— Leveraging Artificial Intelligence to Combat Climate Change: Opportunities, Challenges, and Recommendations.” SSRN Electronic Journal [online], 1–55. Deng, L., Zhong, W., Zhao, L., He, X., Lian, Z., Jiang, S., and Chen, C.Y.C. (2020), “Artificial Intelligence-Based Application to Explore Inhibitors of Neurodegene- rative Diseases.” Frontiers in Neurorobotics, 14, 108. Goh, K.H., Wang, L., Yeow, A.Y.K., Poh, H., Li, K., Yeow, J.J.L., and Tan, G.Y.H.

(2021), “Artificial intelligence in sepsis early prediction and diagnosis using unstructured data in healthcare.” Nature Communications 2021 12:1 [online], 12(1), 1–10. Gurovich, Y., Hanani, Y., Bar, O., Nadav, G., Fleischer, N., Gelbman, D., Basel-Salmon, L., Krawitz, P.M., Kamphausen, S.B., Zenker, M., Bird, L.M., and Gripp, K.W. (2019), “Identifying facial phenotypes of genetic disorders using deep learning.” Nature medicine [online], 25(1), 60–64. Haenssle, H.A., Fink, C., Schneiderbauer, R., Toberer, F., Buhl, T., Blum, A.,

Kalloo, A., Hassen, A.B.H., Thomas, L., Enk, A., and Uhlmann, L. (2018), “Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatolo- gists.” Annals of oncology : official journal of the European Society for Medical Oncology [online], 29(8), 1836–1842. Hannun, A.Y., Rajpurkar, P., Haghpanahi, M., Tison, G.H., Bourn, C., Turakhia, M.P., and Ng, A.Y. (2019), “Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network. Nature medicine [online], 25(1), 65–69. Hie, B., Zhong, E.D., Berger, B., and Bryson, B. (2021), “Learning the language of

viral evolution and escape.” Science [online], 371(6526), 284–288. Kalra, J. (Jay), Rafid-Hamed, Z. and Seitzinger, P. (2021), “Artificial Intelligence and Humanistic Medicine: A Symbiosis.” In: Kalra J., Lightner N.J., Taiar R. (eds). Advances in Human Factors and Ergonomics in Healthcare and Medical Devices. AHFE 2021. Lecture Notes in Networks and Systems. Cham: Springer, 263. 3–8. Kalra, J., Rafid-Hamed, Z., and Seitzinger, P. (2019), “The Future of Diagnostic

Laboratory Testing in Healthcare.” American Journal of Biomedical Science & Research, 5(2), 89–91. Liang, H., Tsui, B.Y., Ni, H., Valentim, C., Baxter, S.L., Liu, G., Cai, W., Kermany, D.S., Sun, X., Chen, J., and He, L. (2019), “Evaluation and accurate dia- gnoses of pediatric diseases using artificial intelligence. Nature Medicine [online], 25(3), 433–438.

156 Seitzinger and Kalra

Paiva, O.A. and Prevedello, L.M. (2017), “The potential impact of artificial intelligence in radiology.” Radiologia Brasileira [online], 50(5), 5–6. Plowright, R.K. and Hudson, P.J. (2021), “From protein to pandemic: The transdi-

sciplinary approach needed to prevent spillover and the next pandemic.” Viruses [online], 13(7), 1298. Schuman, A.J. (2014), “Speed EHR documentation with voice recognition software: Incorporating voice recognition technology into your EHR will help you spend less time documenting and more time with patients.” Contemporary Pediatrics [online], 31(6), 28, 30–32. Seitzinger, P., Rafid-Hamed, Z., and Kalra, J., (2021), “Diagnostic Decision Making

in the Face of Uncertainty: COVID-19 and Its Sequalae.” In: Kalra J., Lightner N.J., Taiar R. (eds). Advances in Human Factors and Ergonomics in Health- care and Medical Devices. AHFE 2021. Lecture Notes in Networks and Systems. Cham: Springer, 263. 763–769. Seitzinger, P., Rafid-Hamed, Z., and Kalra, J., (2021), “Global Trends in Clinical Practice and Healthcare Delivery: Opportunities for Growth and Innovation.” In: Kalra J., Lightner N.J., Taiar R. (eds). Advances in Human Factors and Ergono- mics in Healthcare and Medical Devices. AHFE 2021. Lecture Notes in Networks and Systems. Cham: Springer, 263. 9–15. Tajmir, S.H., Lee, H., Shailam, R., Gale, H.I., Nguyen, J.C., Westra, S.J., Lim, R.,

Yune, S., Gee, M.S., and Do, S. (2019), “Artificial intelligence-assisted inter- pretation of bone age radiographs improves accuracy and decreases variability.” Skeletal radiology [online], 48(2), 275–283. Wang, J., Ju, R., Chen, Y., Zhang, L., Hu, J., Wu, Y., Dong, W., Zhong, J., and Yi, Z. (2018), “Automated retinopathy of prematurity screening using deep neural networks.” EBioMedicine [online], 35, 361–368.

Cover of the proceedings volume Healthcare and Medical Devices

Explore the complete digital volume

Healthcare and Medical Devices

Read this individual research paper free through the original publisher, or explore the full digital proceedings book.

ISBN 978-1-958651-27-8

Explore this ISBN in the AHFE Bookstore Listed in the July 2026 bookstore catalogue; search by ISBN to confirm current availability · Digital book purchase optional

Explore Related Research

Related research 8 matches

More papers from this volume

← Back to Research ExplorerBrowse the full library →