Learning meaningful latent space representations for patient risk stratification: Model development and validation for dengue and other acute febrile illness.

Autor:	Hernandez B; Centre for Bio-Inspired Technology, Imperial College London, London, United Kingdom.; Centre for Amtimicrobial Optimisation, Imperial College London, London, United Kingdom., Stiff O; Centre for Bio-Inspired Technology, Imperial College London, London, United Kingdom., Ming DK; Centre for Amtimicrobial Optimisation, Imperial College London, London, United Kingdom.; NIHR HPRU in Healthcare Associated Infections and Antimicrobial Resistance, Imperial College London, London, United Kingdom., Ho Quang C; Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam., Nguyen Lam V; Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam.; University of Medicine and Pharmacy, Ho Chi Minh City, Vietnam., Nguyen Minh T; Children's Hospital No 1, Ho Chi Minh City, Vietnam., Nguyen Van Vinh C; Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam.; Hospital for Tropical Diseases, Ho Chi Minh City, Vietnam., Nguyen Minh N; Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam., Nguyen Quang H; Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam., Phung Khanh L; Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam.; University of Medicine and Pharmacy, Ho Chi Minh City, Vietnam., Dong Thi Hoai T; Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam., Dinh The T; Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam., Huynh Trung T; Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam.; Hospital for Tropical Diseases, Ho Chi Minh City, Vietnam., Wills B; Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam.; Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, Oxford, United Kingdom., Simmons CP; Institute of Vector Borne Disease, Monash University, Melbourne, VIC, Australia., Holmes AH; Centre for Amtimicrobial Optimisation, Imperial College London, London, United Kingdom.; NIHR HPRU in Healthcare Associated Infections and Antimicrobial Resistance, Imperial College London, London, United Kingdom., Yacoub S; Oxford University Clinical Research Unit, Ho Chi Minh City, Vietnam.; Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, Oxford, United Kingdom., Georgiou P; Centre for Bio-Inspired Technology, Imperial College London, London, United Kingdom.; Centre for Amtimicrobial Optimisation, Imperial College London, London, United Kingdom.
Jazyk:	angličtina
Zdroj:	Frontiers in digital health [Front Digit Health] 2023 Feb 22; Vol. 5, pp. 1057467. Date of Electronic Publication: 2023 Feb 22 (Print Publication: 2023).
DOI:	10.3389/fdgth.2023.1057467
Abstrakt:	Background: Increased data availability has prompted the creation of clinical decision support systems. These systems utilise clinical information to enhance health care provision, both to predict the likelihood of specific clinical outcomes or evaluate the risk of further complications. However, their adoption remains low due to concerns regarding the quality of recommendations, and a lack of clarity on how results are best obtained and presented. Methods: We used autoencoders capable of reducing the dimensionality of complex datasets in order to produce a 2D representation denoted as latent space to support understanding of complex clinical data. In this output, meaningful representations of individual patient profiles are spatially mapped in an unsupervised manner according to their input clinical parameters. This technique was then applied to a large real-world clinical dataset of over 12,000 patients with an illness compatible with dengue infection in Ho Chi Minh City, Vietnam between 1999 and 2021. Dengue is a systemic viral disease which exerts significant health and economic burden worldwide, and up to 5% of hospitalised patients develop life-threatening complications. Results: The latent space produced by the selected autoencoder aligns with established clinical characteristics exhibited by patients with dengue infection, as well as features of disease progression. Similar clinical phenotypes are represented close to each other in the latent space and clustered according to outcomes broadly described by the World Health Organisation dengue guidelines. Balancing distance metrics and density metrics produced results covering most of the latent space, and improved visualisation whilst preserving utility, with similar patients grouped closer together. In this case, this balance is achieved by using the sigmoid activation function and one hidden layer with three neurons, in addition to the latent dimension layer, which produces the output (Pearson, 0.840; Spearman, 0.830; Procrustes, 0.301; GMM 0.321). Conclusion: This study demonstrates that when adequately configured, autoencoders can produce two-dimensional representations of a complex dataset that conserve the distance relationship between points. The output visualisation groups patients with clinically relevant features closely together and inherently supports user interpretability. Work is underway to incorporate these findings into an electronic clinical decision support system to guide individual patient management. Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer [ST] declared a shared affiliation with the author [CPS] to the handling editor at the time of review. (© 2023 Hernandez, Stiff, Ming, Ho Quang, Nguyen Lam, Nguyen Minh, Nguyen Van Vinh, Nguyen Minh, Nguyen Quang, Phung Khanh, Dong Thi Hoai, Dinh The, Huynh Trung, Wills, Simmons, Holmes, Yacoub and Georgiou.)
Databáze:	MEDLINE
Externí odkaz:	Zobrazit plný text záznamu