tetano
Editor, Senior Moderator
Sci Rep
. 2025 Feb 7;15(1):4669.
doi: 10.1038/s41598-025-88976-4. Detection of serum composition in pediatric inflammatory multisystem syndrome associated with SARS-CoV-2 and the response for the treatment by FTIR
Joanna Depciuch[SUP] 1 2 [/SUP], Izabela Sieminska[SUP] 3 4 [/SUP], Pawel Jakubczyk[SUP] 5 [/SUP], Bartosz Klebowski[SUP] 6 [/SUP], Katarzyna Ptak[SUP] 7 [/SUP], Izabela Szymońska[SUP] 7 [/SUP], Przemko Kwinta[SUP] 7 8 [/SUP], Maciej Siedlar[SUP] 4 9 [/SUP], Jan Jakub Kęsik[SUP] 10 [/SUP], Magdalena Parlinska-Wojtan[SUP] 6 [/SUP], Jarek Baran[SUP] 4 9 [/SUP]
Affiliations
Pediatric Inflammatory Multisystem Syndrome (PIMS-TS), associated with SARS-CoV-2 infection, is a severe complication after COVID-19 in children. It is caused by the immune reaction to SARS-CoV-2, and usually appears three to six weeks after the infection. Unfortunately, PIMS causes non-specific symptoms, which makes its diagnosis and treatment difficult. In this paper, we propose Fourier Transform InfraRed spectrometry (FTIR) to identify chemical changes in blood serum of children induced by PIMS and caused by subsequent treatment of the syndrome. The results suggest that although the Principal Component Analysis (PCA) of FTIR data did not allow for differentiation of healthy children and children with PIMS before and after the treatment, the implementation of Support Vector Machine (SVM) showed that the accuracy of the FTIR region between 800 cm[SUP]- 1[/SUP] and 1800 cm[SUP]- 1[/SUP] in PIMS detection is as high as 92% with a sensitivity of 100%. The difference in the chemical compositions of sera from the control group and the children after the treatment was detected in 54%, indicating that the treatment was effective. Indeed, the obtained medical data clearly showed a decrease of C-reactive protein (CRP) and Procalcitonin (PCT) concentration in serum after the treatment. The decision tree showed that peak 1455 cm[SUP]- 1[/SUP] could be used as a potential FTIR PIMS marker. Importantly, FTIR data correlates well with medical parameters, however the correlation differs with respect to the groups before and after the treatment.
Keywords: FTIR; Liquid biopsy; Machine learning; PIMS; Serum marker.
. 2025 Feb 7;15(1):4669.
doi: 10.1038/s41598-025-88976-4. Detection of serum composition in pediatric inflammatory multisystem syndrome associated with SARS-CoV-2 and the response for the treatment by FTIR
Joanna Depciuch[SUP] 1 2 [/SUP], Izabela Sieminska[SUP] 3 4 [/SUP], Pawel Jakubczyk[SUP] 5 [/SUP], Bartosz Klebowski[SUP] 6 [/SUP], Katarzyna Ptak[SUP] 7 [/SUP], Izabela Szymońska[SUP] 7 [/SUP], Przemko Kwinta[SUP] 7 8 [/SUP], Maciej Siedlar[SUP] 4 9 [/SUP], Jan Jakub Kęsik[SUP] 10 [/SUP], Magdalena Parlinska-Wojtan[SUP] 6 [/SUP], Jarek Baran[SUP] 4 9 [/SUP]
Affiliations
- PMID: 39920293
- PMCID: PMC11805954
- DOI: 10.1038/s41598-025-88976-4
Pediatric Inflammatory Multisystem Syndrome (PIMS-TS), associated with SARS-CoV-2 infection, is a severe complication after COVID-19 in children. It is caused by the immune reaction to SARS-CoV-2, and usually appears three to six weeks after the infection. Unfortunately, PIMS causes non-specific symptoms, which makes its diagnosis and treatment difficult. In this paper, we propose Fourier Transform InfraRed spectrometry (FTIR) to identify chemical changes in blood serum of children induced by PIMS and caused by subsequent treatment of the syndrome. The results suggest that although the Principal Component Analysis (PCA) of FTIR data did not allow for differentiation of healthy children and children with PIMS before and after the treatment, the implementation of Support Vector Machine (SVM) showed that the accuracy of the FTIR region between 800 cm[SUP]- 1[/SUP] and 1800 cm[SUP]- 1[/SUP] in PIMS detection is as high as 92% with a sensitivity of 100%. The difference in the chemical compositions of sera from the control group and the children after the treatment was detected in 54%, indicating that the treatment was effective. Indeed, the obtained medical data clearly showed a decrease of C-reactive protein (CRP) and Procalcitonin (PCT) concentration in serum after the treatment. The decision tree showed that peak 1455 cm[SUP]- 1[/SUP] could be used as a potential FTIR PIMS marker. Importantly, FTIR data correlates well with medical parameters, however the correlation differs with respect to the groups before and after the treatment.
Keywords: FTIR; Liquid biopsy; Machine learning; PIMS; Serum marker.