Advances of Photothermal Agents with Fluorescence Imaging/Enhancement Ability in the Field of Photothermal Therapy and Diagnosis.

Photothermal therapy (PTT) is an effective cancer treatment method. Due to its easy focusing and tunability of the irradiation light, direct and accurate local treatment can be performed in a noninvasive manner by PTT. This treatment strategy requires the use of photothermal agents to convert light energy into heat energy, thereby achieving local heating and triggering biochemical processes to kill tumor cells. As a key factor in PTT, the photothermal conversion ability of photothermal agents directly determines the efficacy of PTT. In addition, photothermal agents generally have photothermal imaging (PTI) and photoacoustic imaging (PAI) functions, which can not only guide the optimization of irradiation conditions but also achieve the integration of disease diagnosis. If the photothermal agents have function of fluorescence imaging (FLI) or fluorescence enhancement, they can not only further improve the accuracy in disease diagnosis but also accurately determine the tumor location through multimodal imaging for corresponding treatment. In this paper, we summarize recent advances in photothermal agents with FLI or fluorescence enhancement functions for PTT and tumor diagnosis. According to the different recognition sites, the application of specific targeting photothermal agents is introduced. Finally, limitations and challenges of photothermal agents with fluorescence imaging/enhancement in the field of PTT and tumor diagnosis are prospected.

Maintaining moderate versus lower PEEP after cardiac surgery: a propensity-scored matched analysis.

BACKGROUND: Setting positive end-expiratory pressure (PEEP) at around 5 cm H2O in the early postoperative period seems a common practice for most patients. It remains unclear if the routine application of higher levels of PEEP confers any meaningful clinical benefit for cardiac surgical patients. The aim of this study was to compare moderate versus conventional lower PEEP on patient-centered outcomes in the intensive care unit (ICU). METHODS: This is a single-center retrospective study involving patients receiving cardiac surgery from June 2022 to May 2023. Propensity-score matching (PSM) was used to balance the baseline differences. Primary outcomes were the duration of mechanical ventilation and ICU length of stay. Secondary outcomes included PaO2/FiO2 ratio at 24 h and the need for prone positioning during ICU stay. RESULTS: A total of 334 patients were included in the study, 102 (31%) of them received moderate PEEP (≥ 7 cm H2O) for the major time in the early postoperative period (12 h). After PSM, 79 pairs of patients were matched with balanced baseline data. The results showed that there was marginal difference in the distribution of mechanical ventilation duration (p = 0.05) and the Moderate PEEP group had a higher extubation rate at the day of T-piece trial (65 [82.3%] vs 52 [65.8%], p = 0.029). Applying moderate PEEP was also associated with better oxygenation. No differences were found regarding ICU length of stay and patients requiring prone positioning between groups. CONCLUSION: In selective cardiac surgical patients, using moderate PEEP compared with conventional lower PEEP in the early postoperative period correlated to better oxygenation, which may have potential for earlier liberation of mechanical ventilation.

Sublingual microcirculation: comparison between the 415 nm blue light and 520 nm green light of sidestream dark field videomicroscopes.

Green light with a wavelength of 520 nm is commonly used in sidestream dark field (SDF) video microscopes for sublingual microcirculation assessment in clinical practice. However, blue light could obtain a clearer microcirculatory image due to a higher light absorption coefficient of hemoglobin. The aim of this study was to compare the sublingual microcirculatory image quality acquisition and related microcirculatory parameters between 520 nm green light and 415 nm blue light probes in the SDF device named MicroSee V100. Sublingual microcirculation films from twenty-one healthy volunteers were prospectively collected by blue light and green light probes, and only one video of each wavelength was recorded and analyzed in each volunteer. Moreover, 200 sublingual microcirculation films (100 by blue light probe and 100 by green light probe) of ICU patients were retrospectively scored for microcirculation image quality. Compared to green light, an increase in the perfused vessel density (paired t test, increased by 4.6 ± 4.7 mm/mm2, P < 0.0001) and total vessel density (paired t test, increased by 5.1 ± 4.6 mm/mm2, P < 0.0001) was observed by blue light in the healthy volunteers. The blue light probe had a significantly lower rate of unacceptable films than the green light probe in the 200 films of ICU patients (10/100 vs. 39/100, P < 0.0001). Blue light provides a higher microcirculatory vessel density and image quality than the existing SDF probe using green light.

Bedside electrical impedance tomography in early diagnosis of pneumothorax in mechanically ventilated ICU patients - a single-center retrospective cohort study.

PURPOSE: This study aimed to evaluate the routine use of electrical impedance tomography (EIT) to diagnose pneumothorax (PTX) in mechanically ventilated patients in the intensive care unit (ICU). METHODS: A retrospective cohort study was conducted including mechanically ventilated supine patients who received EIT examinations. The EIT-based tidal variation was divided into ventral and dorsal regions of interest (ROIs): upper right (UR, ROI1), upper left (UL, ROI2) lower right (LR, ROI3), and lower left (LL, ROI4), and the ventilation defect score (DS) was calculated in each quadrant. Furthermore, horizontal ventral ventilation index (HVVI) was defined as ROI1% / ROI2% in the two ventral quadrants if ROI1% > ROI2%, otherwise HVVI = ROI2% / ROI1%. RESULTS: A total of 203 patients were included, 25 of them with confirmed PTX. In the PTX patients, preceding cardiac surgery was the most common cause of PTX. Compared with the patients without PTX, the PTX patients had a higher DS in the ventral quadrants [median and interquartile range (IQR): 1.00 (0.00, 2.00) vs. 0.00 (0.00, 0.00), P < 0.001] respectively, but similar in the dorsal quadrants [median and IQR: 1.00 (0.00, 1.00) vs. 0.00 (0.00, 1.00), P = 0.722]. Moreover, a higher HVVI was found in the PTX group [median and IQR: 2.51 (1.58, 3.52) vs. 1.36 (1.15, 1.77), P < 0.001]. The area under the receiver operating characteristic curve of the HVVI to differentiate PTX from non-PTX was 0.88, with a sensitivity of 70% and a specificity of 90% when the cut-off value was 2.57. CONCLUSION: The ventilation defect in the ventral regions and a high HVVI on EIT were observed in mechanically ventilated patients with PTX, which should trigger further diagnostics to confirm it.

The relationship between ventilatory ratio (VR) and 28-day hospital mortality by restricted cubic splines (RCS) in 14,328 mechanically ventilated ICU patients.

BACKGROUND: Previous studies found that high levels of ventilatory ratio (VR) were associated with a poor prognosis due to worse ventilatory efficiency in acute respiratory distress syndrome patients. However, relatively few large studies have assessed the association between VR and intensive care unit (ICU) mortality in the general adult ventilated population. METHODS: The present study is a retrospective cohort study. Patients mechanically ventilated for more than 12 h were included. VR was calculated based on a previously reported formula. Restricted cubic spline models were used to fit the relationship between VR and mortality risks. RESULTS: A total of 14,328 mechanically ventilated ICU patients were included in the study, of which 1311 died within 28 days. The results of the study are as follows: (1) In the general adult ventilated population, VR was positively associated with 28-day mortality when VR ≥ 1.3 (increase of 0.1 per VR; HR 1.05, p < 0.001). The same tendency was also observed in the populations of severe hypoxemia with a PaO2/FiO2 (P/F) ratio < 200 mmHg. (2) However, in the population with a P/F ratio ≥ 200, a J-shaped dose-response association between VR and the risk of mortality was observed, with the risk of death positively associated with VR when VR ≥ 0.9 (10% increase in HR for every 0.1 increase in VR, p = 0.000) but negatively associated with VR when VR < 0.9 (10% decrease in HR for every 0.1 increase in VR, p = 0.034). In the population of P/F ratio ≥ 200 with VR less than 0.9, compared to the survival group, the nonsurvival group had a lower level PCO2 (33 mmHg [29.1, 37.9] vs. 34.4 mmHg [30.6, 38.5]), rather than a significant level of measured minute ventilation or P/F ratio. CONCLUSIONS: VR was positively associated with the risk of death in the general ICU population; however, VR was inversely associated with 28-day mortality in the population with a P/F ratio ≥ 200 and low VR . Further research should investigate this relationship, and VR should be interpreted with caution in clinical practice.

Early individualized positive end-expiratory pressure guided by electrical impedance tomography in acute respiratory distress syndrome: a randomized controlled clinical trial.

BACKGROUND: Individualized positive end-expiratory pressure (PEEP) by electrical impedance tomography (EIT) has potential interest in the optimization of ventilation distribution in acute respiratory distress syndrome (ARDS). The aim of the study was to determine whether early individualized titration of PEEP with EIT improved outcomes in patients with ARDS. METHODS: A total of 117 ARDS patients receiving mechanical ventilation were randomly assigned to EIT group (n = 61, PEEP adjusted based on ventilation distribution) or control group (n = 56, low PEEP/FiO2 table). The primary outcome was 28-day mortality. Secondary and exploratory outcomes were ventilator-free days, length of ICU stay, incidence of pneumothorax and barotrauma, and difference in Sequential Organ Failure Assessment (SOFA) score at day 1 (ΔD1-SOFA) and day 2 (ΔD2-SOFA) compared with baseline. MEASUREMENTS AND MAIN RESULTS: There was no statistical difference in the value of PEEP between the EIT group and control group, but the combination of PEEP and FiO2 was different between groups. In the control group, a significantly positive correlation was found between the PEEP value and the corresponding FiO2 (r = 0.47, p < 0.00001) since a given matched table was used for PEEP settings. Diverse combinations of PEEP and FiO2 were found in the EIT group (r = 0.05, p = 0.68). There was no significant difference in mortality rate (21% vs. 27%, EIT vs. control, p = 0.63), ICU length of stay (13.0 (7.0, 25.0) vs 10.0 (7.0, 14.8), median (25th-75th percentile); p = 0.17), and ventilator-free days at day 28 (14.0 (2.0, 23.0) vs 19.0 (0.0, 24.0), p = 0.55) between the two groups. The incidence of new barotrauma was zero. Compared with control group, significantly lower ΔD1-SOFA and ΔD2-SOFA were found in the EIT group (p < 0.001) in a post hoc comparison. Moreover, the EIT group exhibited a significant decrease of SOFA at day 2 compared with baseline (paired t-test, difference by - 1 (- 3.5, 0), p = 0.001). However, the control group did show a similar decrease (difference by 1 (- 2, 2), p = 0.131). CONCLUSION: Our study showed a 6% absolute decrease in mortality in the EIT group: a statistically non-significant, but clinically non-negligible result. This result along with the showed improvement in organ function might justify further reserach to validate the beneficial effect of individualized EIT-guided PEEP setting on clinical outcomes of patients with ARDS. TRIAL REGISTRATION: ClinicalTrials, NCT02361398. Registered 11 February 2015-prospectively registered, https://clinicaltrials.gov/show/NCT02361398 .

Twenty-four-hour mechanical power variation rate is associated with mortality among critically ill patients with acute respiratory failure: a retrospective cohort study.

OBJECTIVES: Defined as the energy applied to the respiratory system by ventilator, mechanical power (MP) of ventilation reflects the risk of ventilation-induced lung injury. This study aims to explore the relationship between dynamic changes in MP and prognosis in critically ill patients. METHODS: This was a single-centre retrospective cohort study. Patients receiving mechanical ventilation with acute respiratory failure (ARF) and MP > 10 J/min on admission in the ICU were included. MP (J/min) was calculated as 0.098 × minute ventilation (L/min) × [(peak inspiratory pressure + positive end-expiratory pressure)/2] and the MP variation rate (%) as ([baseline MP - 24-h MP]/baseline MP) × 100. Patients were divided into two groups according to whether MP decreased 24 h after admission (MP-improved group defined as 24-h MP variation rate > 0% vs. MP-worsened group defined as 24-h MP variation rate ≤ 0%). RESULTS: In total, 14,463 patients were screened between January 2015 and June 2020, and finally, a study cohort of 602 patients was obtained. The MP-improved group had a lower ICU mortality rate than the MP-worsened group (24% vs. 36%; p = 0.005). The 24-h MP variation rate was associated with ICU mortality after adjusting for confounders (odds ratio, 0.906 [95% CI 0.833-0.985]; p = 0.021), while baseline MP (p = 0.909) and 24-h MP (p = 0.059) were not. All MP components improved in the MP-improved group, while minute ventilation and positive end-expiratory pressure contributed to the increase in MP in the MP-worsened group. CONCLUSIONS: The 24-h MP variation rate was an independent risk factor for ICU mortality among ARF patients with elevated MP. Early decreases in MP may provide prognostic benefits in this population.

Influence of overdistension/recruitment induced by high positive end-expiratory pressure on ventilation-perfusion matching assessed by electrical impedance tomography with saline bolus.

BACKGROUND: High positive end-expiratory pressures (PEEP) may induce overdistension/recruitment and affect ventilation-perfusion matching (VQMatch) in mechanically ventilated patients. This study aimed to investigate the association between PEEP-induced lung overdistension/recruitment and VQMatch by electrical impedance tomography (EIT). METHODS: The study was conducted prospectively on 30 adult mechanically ventilated patients: 18/30 with ARDS and 12/30 with high risk for ARDS. EIT measurements were performed at zero end-expiratory pressures (ZEEP) and subsequently at high (12-15 cmH2O) PEEP. The number of overdistended pixels over the number of recruited pixels (O/R ratio) was calculated, and the patients were divided into low O/R (O/R ratio < 15%) and high O/R groups (O/R ratio ≥ 15%). The global inhomogeneity (GI) index was calculated to evaluate the ventilation distribution. Lung perfusion image was calculated from the EIT impedance-time curves caused by 10 ml 10% NaCl injection during a respiratory pause (> 8 s). DeadSpace%, Shunt%, and VQMatch% were calculated based on lung EIT perfusion and ventilation images. RESULTS: Increasing PEEP resulted in recruitment mainly in dorsal regions and overdistension mainly in ventral regions. ΔVQMatch% (VQMatch% at high PEEP minus that at ZEEP) was significantly correlated with recruited pixels (r = 0.468, P = 0.009), overdistended pixels (r = - 0.666, P < 0.001), O/R ratio (r = - 0.686, P < 0.001), and ΔSpO2 (r = 0.440, P = 0.015). Patients in the low O/R ratio group (14/30) had significantly higher Shunt% and lower VQMatch% than those in the high O/R ratio group (16/30) at ZEEP but not at high PEEP. Comparable DeadSpace% was found in both groups. A high PEEP caused a significant improvement of VQMatch%, DeadSpace%, Shunt%, and GI in the low O/R ratio group, but not in the high O/R ratio group. Using O/R ratio of 15% resulted in a sensitivity of 81% and a specificity of 100% for an increase of VQMatch% > 20% in response to high PEEP. CONCLUSIONS: Change of ventilation-perfusion matching was associated with regional overdistention and recruitment induced by PEEP. A low O/R ratio induced by high PEEP might indicate a more homogeneous ventilation and improvement of VQMatch. TRIAL REGISTRATION: ClinicalTrials.gov, NCT04081155 . Registered on 9 September 2019-retrospectively registered.

Effect of postextubation high-flow nasal cannula therapy on lung recruitment and overdistension in high-risk patient.

BACKGROUND: Postextubation high-flow nasal cannula (HFNC) is used as a support therapy in high-risk patients in ICU. This study aimed to determine the effects of HFNC therapy on lung recruitment and overdistension assessed by electrical impedance tomography (EIT). METHODS: Twenty-four patients who received HFNC within 24 h after extubation were prospectively enrolled in this study. EIT was used to monitor regional lung ventilation distributions at baseline (conventional oxygen therapy) and three flow rate levels of HFNC therapy (20, 40, and 60 L/min). Change of end-expiratory lung impedance (ΔEELI), regional recruitment (recruited-pixels) and overdistension (overdistended-pixels), and lung strain change were determined by EIT. EIT images were equally divided into four ventral-to-dorsal horizontal regions of interest (ROIs 1, 2, 3, and 4). "Overdistension-by HFNC" due to HFNC is defined as an increase of overdistened-pixels > 10 than baseline. Patients were divided into two groups: (1) high potential of recruitment (HPR), recruited-pixels > 10 pixels at 60 L/min than baseline, and (2) low potential of recruitment (LPR), recruited-pixels < 10 pixels at 60 L/min than baseline. RESULTS: When the flow rate gradually increased from baseline to 60 L/min, a significant and consistent increasing trend of global ΔEELI (%) (p < 0.0001), recruited-pixels (p < 0.001), and overdistended-pixels (p = 0.101) was observed. Moreover, the increase of ΔEELI was mainly distributed in ROI2 (p = 0.001) and ROI3 (p < 0.0001). The HPR group (13/24 patients) had significantly higher recruited-pixels than the LPR group (11/24 patients) at 20, 40, and 60 L/min. There were no significant differences in PaO2/FiO2, ΔEELI (%), and overdistention pixels between the two groups. The HPR group had 13 patients in which no one had "overdistension-by HFNC", and the LPR group had 11 patients in which 4 patients had "overdistension-by HFNC" (0/13 vs. 4/11, p = 0.017). CONCLUSIONS: Using EIT could identify diverse effects of HFNC on lung regional ventilation in postextubation situations. Further study is required to validate using "HFNC effect" based on lung recruitment and overdistension by EIT in clinical practice. TRIAL REGISTRATION: The study was retrospectively registered at www.clinicaltrials.gov (no. NCT04245241).

Three patterns of osteoarticular involvement in SAPHO syndrome: a cluster analysis based on whole body bone scintigraphy of 157 patients.

OBJECTIVES: To explore the patterns of osteoarticular involvement in SAPHO syndrome. METHODS: Baseline clinical characteristics and imaging data of 99mTc-MDP whole-body bone scintigraphy (WBBS) were collected from 157 out of 164 patients diagnosed with SAPHO syndrome. The twelve most frequently involved osteoarticular sites were analysed by hierarchical cluster analysis with the Ward minimum-variance method. RESULTS: Three distinctive patterns of osteoarticular involvement were identified: the spinal type (70 patients, 44.6%), with predominantly thoracic, lumbar or sacral vertebral lesions; the costal type (52 patients, 33.1%), with lesions of anterior ribs, particularly the first ribs; and the sternoclavicular type (35 patients, 22.3%), with predominantly sternal and bilateral sternoclavicular lesions, characterized by the typical bullhead sign. Notably, a total of 77 (49%) patients exhibited lesions of ribs on WBBS, of which 61.3% involved the first ribs. Interestingly, patients of spinal type were older at onset of cutaneous manifestations than those of sternoclavicular type (P = 0.036) and costal type (P = 0.035). The disease course was remarkably longer in sternoclavicular type than costal type (P = 0.001) and spinal type (P < 0.001). CONCLUSION: The osteoarticular involvement in SAPHO syndrome can be categorized as three distinct patterns with different corresponding clinical features. The costal involvement in SAPHO syndrome, which was under-recognized previously, may define a distinct sub-type of the disease.

Improvement of Asia-Pacific colorectal screening score and evaluation of its use combined with fecal immunochemical test.

BACKGROUND: The Asia-Pacific Colorectal Screening (APCS) score is effective to screen high-risk groups of advanced colorectal neoplasia (ACN) patients but needs revising and can be combined with the fecal immunochemical test (FIT). This paper aimed to improve the APCS score and evaluate its use with the FIT in stratifying the risk of ACN. METHODS: This prospective and multicenter study enrolled 955 and 1201 asymptomatic Chinese participants to form the derivation and validation set, respectively. Participants received the risk factor questionnaire, colonoscopy and FIT. Multiple logistic regression was applied, and C-statistic, sensitivity and negative predictive values (NPVs) were used to compare the screening efficiency. RESULTS: A modified model was developed incorporating age, body mass index (BMI), family history, diabetes, smoking and drinking as risk factors, stratifying subjects into average risk (AR) or high risk (HR). In the validation set, the HR tier group had a 3.4-fold (95% CI 1.8-6.4) increased risk for ACN. The C-statistic for the modified score was 0.69 ± 0.04, and 0.67 ± 0.04 for the original score. The sensitivity of the modified APCS score combined with FIT for screening ACN high-risk cohorts was 76.7% compared with 36.7% of FIT alone and 70.0% of the modified APCS score alone. The NPVs of the modified score combined with FIT for ACN were 98.0% compared with 97.0% of FIT alone and 97.9% of the modified APCS score alone. CONCLUSIONS: The modified score and its use with the FIT are efficient in selecting the HR group from a Chinese asymptomatic population.

A facile three-step total synthesis of tanshinone I.

A facile synthetic approach for total synthesis of tanshinone I has been accomplished. The key precursor is a novel compound, epoxy phenanthraquinone. And this synthesis of tanshinone I is achieved in only three simple stages, which include Diels-Alder reaction, Δ(2)-Weitz-Scheffer-type epoxidation, and Feist-Bénary reaction from commercially available styrene.

Early bedside detection of pulmonary perfusion defect by electrical impedance tomography after pulmonary endarterectomy.

Pulmonary endarterectomy (PEA) is the standard treatment for chronic thromboembolic pulmonary hypertension. However, it poses risks of perioperative vascular complications, which can lead to serious clinical outcomes. This study introduces a novel noninvasive and radiation-free clinical imaging tool, electrical impedance tomography (EIT), for real-time bedside assessment of lung perfusion after PEA. It identifies ventilation-perfusion mismatches arising from postoperative complications, particularly valuable for patients with hemodynamic instability, thus eliminating risks tied to CT room transfers. The article reports a case where EIT was used to identify an in-situ thrombosis post-PEA, marking the first such application. The emphasis is on early detection using EIT, which offers a promising approach for therapeutic interventions and improved postoperative evaluations.

Phenotypes of esophageal pressure response to the change of positive end-expiratory pressure in patients with moderate acute respiratory distress syndrome.

Background: Esophageal pressure (Pes) has been used as a surrogate of pleural pressure (Ppl) to titrate positive end-expiratory pressure (PEEP) in acute respiratory distress syndrome (ARDS) patients. The relationship between Pes and PEEP remains undetermined. Methods: A gastric tube with a balloon catheter was inserted to monitor Pes in moderate to severe ARDS patients who underwent invasive mechanical ventilation. To assess the end-expiratory Pes response (ΔPes) to PEEP changes (ΔPEEP), the PEEP level was decreased and increased subsequently (with an average change of 3 cmH2O). The patients underwent the following two series of PEEP adjustment: (I) from PEEP-3 cmH2O to PEEPbaseline; and (II) from PEEPbaseline to PEEP+3 cmH2O. The patients were classified as "PEEP-dependent type" if they had ΔPes ≥30% ΔPEEP and were otherwise classified as "PEEP-independent type" (ΔPes <30% ΔPEEP in any series). Results: In total, 54 series of PEEP adjustments were performed in 18 ARDS patients. Of these patients, 12 were classified as PEEP-dependent type, and six were classified as PEEP-independent type. During the PEEP adjustment, end-expiratory Pes changed significantly in the PEEP-dependent patients, who had a Pes of 10.8 (7.9, 12.3), 12.5 (10.5, 14.9), and 14.5 (13.1, 18.3) cmH2O at PEEP-3 cmH2O, PEEPbaseline, and PEEP+3 cmH2O, respectively (median and quartiles; P<0.0001), while end-expiratory transpulmonary pressure (PL) was maintained at an optimal range [-0.1 (-0.7, 0.4), 0.1 (-0.6, 0.5), and 0.3 (-0.3, 0.7) cmH2O, respectively]. In the PEEP-independent patients, the Pes remained unchanged, with a Pes of 15.4 (11.4, 17.8), 15.5 (11.6, 17.8), and 15.4 (11.7, 18.30) cmH2O at each of the three PEEP levels, respectively. Meanwhile, end-expiratory PL significantly improved [from -5.5 (-8.5, -3.4) at PEEP-3 cmH2O to -2.5 (-5.0, -1.6) at PEEPbaseline to -0.5 (-1.8, 0.3) at PEEP+3 cmH2O; P<0.01]. Conclusions: Two types of Pes phenotypes were identified according to the ΔPes to ΔPEEP. The underlying mechanisms and implications for clinical practice require further exploration.

A privacy-preserving platform oriented medical healthcare and its application in identifying patients with candidemia.

Federated learning (FL) has emerged as a significant method for developing machine learning models across multiple devices without centralized data collection. Candidemia, a critical but rare disease in ICUs, poses challenges in early detection and treatment. The goal of this study is to develop a privacy-preserving federated learning framework for predicting candidemia in ICU patients. This approach aims to enhance the accuracy of antifungal drug prescriptions and patient outcomes. This study involved the creation of four predictive FL models for candidemia using data from ICU patients across three hospitals in China. The models were designed to prioritize patient privacy while aggregating learnings across different sites. A unique ensemble feature selection strategy was implemented, combining the strengths of XGBoost's feature importance and statistical test p values. This strategy aimed to optimize the selection of relevant features for accurate predictions. The federated learning models demonstrated significant improvements over locally trained models, with a 9% increase in the area under the curve (AUC) and a 24% rise in true positive ratio (TPR). Notably, the FL models excelled in the combined TPR + TNR metric, which is critical for feature selection in candidemia prediction. The ensemble feature selection method proved more efficient than previous approaches, achieving comparable performance. The study successfully developed a set of federated learning models that significantly enhance the prediction of candidemia in ICU patients. By leveraging a novel feature selection method and maintaining patient privacy, the models provide a robust framework for improved clinical decision-making in the treatment of candidemia.