Visual discrimination of tissue during surgery can be challenging since different tissues appear similar to the human eye. Hyperspectral imaging (HSI) removes this limitation by associating each pixel with high-dimensional spectral information. While previous work has shown its general potential to discriminate tissue, clinical translation has been limited due to the method’s current lack of robustness and generalizability. Specifically, the scientific community is lacking a comprehensive spectral tissue atlas, and it is unknown whether variability in spectral reflectance is primarily explained by tissue type rather than the recorded individual or specific acquisition conditions. The contribution of this work is threefold: (1) Based on an annotated medical HSI data set (9059 images from 46 pigs), we present a tissue atlas featuring spectral fingerprints of 20 different porcine organs and tissue types. (2) Using the principle of mixed model analysis, we show that the greatest source of variability related to HSI images is the organ under observation. (3) We show that HSI-based fully-automatic tissue differentiation of 20 organ classes with deep neural networks is possible with high accuracy (> 95%). We conclude from our study that automatic tissue discrimination based on HSI data is feasible and could thus aid in intraoperative decisionmaking and pave the way for context-aware computer-assisted surgery systems and autonomous robotics.

Background: The ultimate goal of haemodynamic therapy is to improve microcirculatory tissue and organ perfusion. Hyperspectral imaging (HSI) has the potential to enable noninvasive microcirculatory monitoring at bedside. Methods: HSI (Tivita^® Tissue System) measurements of tissue oxygenation, haemoglobin, and water content in the skin (ear) and kidney were evaluated in a double-hit porcine model of major abdominal surgery and haemorrhagic shock. Animals of the control group (n = 7) did not receive any resuscitation regime. The interventional groups were treated exclusively with either crystalloid (n = 8) or continuous norepinephrine infusion (n = 7). Results: Haemorrhagic shock led to a drop in tissue oxygenation parameters in all groups. These correlated with established indirect markers of tissue oxygenation. Fluid therapy restored tissue oxygenation parameters. Skin and kidney measurements correlated well. High dose norepinephrine therapy deteriorated tissue oxygenation. Tissue water content increased both in the skin and the kidney in response to fluid therapy. Conclusions: HSI detected dynamic changes in tissue oxygenation and perfusion quality during shock and was able to indicate resuscitation effectivity. The observed correlation between HSI skin and kidney measurements may offer an estimation of organ oxygenation impairment from skin monitoring. HSI microcirculatory monitoring could open up new opportunities for the guidance of haemodynamic management.

Zusammenfassung: Hintergrund: Sepsis ist eine akut lebensbedrohliche multiple Organdysfunktion, ausgelöst durch eine Infektion. Methodik: Bei der vorliegenden Leitlinie handelt es sich um ein Update der S3-Leitlinie „Sepsis – Prävention, Diagnose, Therapie und Nachsorge“ (AMWF-Register-Nr.: 079–001) der Deutschen Sepsis-Gesellschaft (DSG) vom 31.12.2018. Dabei wurde das Update der „Surviving sepsis campaign (SSC): international guidelines for management of sepsis and septic shock 2021“ vom 04.10.2021 als Referenzleitlinie zugrunde gelegt. Die DSG-Leitlinienkommission verglich jede Empfehlung zu den zugrunde liegenden PICO-Fragen der DSG-Leitlinie 2018 (Literaturrecherche bis 12/2018) mit denen der SSC-Leitlinie 2021 (Literaturrecherche bis 07/2019) und bewertete die in der Zwischenzeit neu verfügbare publizierte Datenlage (Literaturrecherche bis 12/2024) mittels systematischer Aktualisierungsrecherchen und Literaturbewertungen unter Befolgung des Regelwerkes des GRADE-Systems und der AWMF. Ergebnisse: Insgesamt wurden 88 PICO-Fragen u. a. zur Diagnose und Therapie der Infektion und des Organversagens adressiert. Davon wurden 2 als Statements, 29 als Expertenkonsens und 57 als evidenzbasierte Empfehlungen (26 mit starkem und 31 mit schwachem Empfehlungsgrad) konsentiert. Im Vergleich zur Vorgänger-Leitlinie 2018 wurden 43 Empfehlungen überprüft, aber beibehalten. 16 Empfehlungen wurden geändert, und 29 Empfehlungen wurden neu ausgesprochen. Schlussfolgerung: Angesichts fehlender Evidenz für zahlreiche Maßnahmen zur stationären Versorgung von Patienten mit Sepsis oder septischem Schock wurden alte und neue Wissenslücken offenbart. Bei den evidenzbasierten Empfehlungen war die zugrunde liegende Evidenzqualität nach GRADE nur bei 5 Empfehlungen hoch, bei 18 Empfehlungen moderat, bei 17 Empfehlungen niedrig und bei 16 sehr niedrig. Diese Evidenzlücken können nur durch zukünftige multizentrische, nichtkommerzielle klinische Prüfungen geschlossen werden. Das Update der S3-Leitlinie Sepsis beinhaltet einige Aktualisierungen zu Empfehlungen der Vorgängerleitlinie. Diese Aktualisierungen werden in einige der fall- und einrichtungsbezogenen QS-Indikatoren des QS-Verfahrens 2025 einfließen müssen. Beeinträchtigungen in der gesundheitsbezogenen Lebensqualität müssen bei Überlebenden mehr in den Fokus der ambulanten Versorgung gerückt werden.

With sepsis remaining a leading cause of mortality, early identification of patients with sepsis and those at high risk of death is a challenge of high socioeconomic importance. Given the potential of hyperspectral imaging (HSI) to monitor microcirculatory alterations, we propose a deep learning approach to automated sepsis diagnosis and mortality prediction using a single HSI cube acquired within seconds. In a prospective observational study, we collected HSI data from the palms and fingers of more than 480 intensive care unit patients. Neural networks applied to HSI measurements predicted sepsis and mortality with areas under the receiver operating characteristic curve (AUROCs) of 0.80 and 0.72, respectively. Performance improved substantially with additional clinical data, reaching AUROCs of 0.94 for sepsis and 0.83 for mortality. We conclude that deep learning–based HSI analysis enables rapid and noninvasive prediction of sepsis and mortality, with a potential clinical value for enhancing diagnosis and treatment.

Septic shock substantially alters the pharmacokinetic properties of β-lactams with a subsequently high risk of insufficiently low serum concentrations and treatment failure. Considering their pharmacokinetic (PK)/pharmacodynamic (PD) index, prolonged infusions (PI) of β-lactams extend the time that the unbound fraction of the drug remains above the minimal inhibitory concentration MIC (ft _>MIC) and may improve patient survival. The present study is a monocentric, retrospective before-and-after analysis of septic shock patients treated with β-lactams. Patients of the years 2015–2017 received intermittent bolus application whereas patients of 2017–2020 received PI of β-lactams. The primary outcome was mortality at day 30 and 90 after diagnosis of septic shock. Mortality rates in the PI group were significantly lower on day 30 (PI: 41%, n = 119/290 vs. IB: 54.8%, n = 68/114; p = 0.0097) and day 90 (PI: 47.9%, n = 139/290 vs. IB: 62.9%, n = 78/124; p = 0.005). After propensity-score matching, 30- and 90-day mortality remained lower for the PI group (−10%, p = 0.14). PI was further associated with a reduction in the duration of invasive ventilation and a stronger decrease in SOFA scores within a 14 day-observation period. PI of β-lactams was associated with a significant reduction of mortality in patients with septic shock and may have beneficial effects on invasive ventilation and recovery from sepsis-related organ failure.