Prevention of dialysis hypotension episodes using fuzzy logic control system

Development and validation of AI-based triage support algorithms for prevention of intradialytic hypotension

BACKGROUND

Intradialytic hypotension remains one of the most recurrent complications of dialysis sessions. Inadequate management can lead to adverse outcomes, highlighting the need to develop personalized approaches for the prevention of intradialytic hypotension. Here, we sought to develop and validate two AI-based risk models predicting the occurrence of symptomatic intradialytic hypotension at different time points.

METHODS

The models were built using the XGBoost algorithm and they predict the occurrence of intradialytic hypotension in the next dialysis session and in the next month. The initial dataset, obtained from routinely collected data in the EuCliD® Database, was split to perform model derivation, training and validation. Model performance was evaluated by concordance statistic and calibration charts; the importance of features was assessed with the Shapley Additive Explanation (SHAP) methodology.

RESULTS

The final dataset included 1,249,813 dialysis sessions, and the incidence rate of intradialytic hypotension was 10.07% (95% CI 10.02-10.13). Our models retained good discrimination (AUC around 0.8) and a suitable calibration yielding to the selection of three classification thresholds identifying four distinct risk groups. Variables providing the most significant impact on risk estimates were blood pressure dynamics and other metrics mirroring hemodynamic instability over time.

CONCLUSIONS

Recurrent symptomatic intradialytic hypotension could be reliably and accurately predicted using routinely collected data during dialysis treatment and standard clinical care. Clinical application of these prediction models would allow for personalized risk-based interventions for preventing and managing intradialytic hypotension.

A novel double parameters based biofeedback system to prevent intradialytic hypotension in clinical practice: Case report

Intradialytic hypotension (IDH) still represents the main complication during dialysis treatment. Patient's discomfort, reduced dialysis efficiency, reduced treatment time are only some of the main problems that could derive from the need to stop or temporarily interrupt the treatment because of IDH. Up to now, different types of treatment (HDF, HF, or AFB) have been considered to prevent IDH as well as the use of biofeedback systems (Blood Volume Monitoring, Temperature control, Sodium profiling, Blood Pressure monitoring). Recently a new biofeedback system (Biologic Fusion^®) controlling simultaneously blood pressure (BP) and relative blood volume changes (RBV) has been developed to prevent IDH. This system runs according to fuzzy logic. We describe how this system works and to a better understanding, we report a clinical case of a patient with frequent IDH treated with the use of this system.

Sodium and ultrafiltration profiling in hemodialysis: A long-forgotten issue revisited

Sodium and ultrafiltration profiling are method of dialysis in which dialysate sodium concentration and ultrafiltration rate are altered during the course of the dialysis session. Sodium and ultrafiltration profiling have been used, commonly simultaneously, to improve hemodynamic stability during hemodialysis. Sodium profiling is particularly effective in decreasing the incidence of intradialytic hypotension, while ultrafiltration profiling is suggested to decrease subclinical repeated end organ ischemia during dialysis. However, complications such as increased interdialytic weight gain and thirst due to sodium excess have prevented widespread use of sodium profiling. Evidence suggest that different sodium profiling techniques may lead to different clinical results, and preferring sodium balance neutral sodium profiling may mitigate adverse effects related to sodium overload. However, evidence is lacking on the long-term clinical outcomes of different sodium profiling methods. Optimal method of sodium profiling as well as the utility of sodium/ultrafiltration profiling in routine practice await further clinical investigation.

Updates on hemodialysis techniques with a common denominator: The personalization of the dialytic therapy

Hemodialysis (HD) is a life-saving therapy for patients with end-stage renal disease. In dialyzed patients, the prevalence of multi-morbidity is rising driven by various factors, such as the population aging, the incomplete correction of uremia, and the side effects of the dialysis therapy itself. Each dialyzed patient has their own specific clinical and biochemical problems. It is therefore unthinkable that the same dialysis procedure can be able to meet the needs of every patient on chronic dialysis. We have very sophisticated dialysis machines and different dialysis techniques and procedures beyond conventional HD, such as hemodiafiltration (HDF) with pre- and post-dilution, acetate-free biofiltration (AFB), hemofiltration (HF), and expanded HD. Each of these techniques has its own specific characteristics. To solve some intradialytic clinical issues, such as arterial hypotension and arrhythmias, we have biofeedback systems with automatic regulation of the blood volume, body temperature, arterial pressure, as well as potassium profiling techniques in the dialysis bath. New technical innovations, such as citrate-containing dialysate or heparin-coated membranes, could reduce the risk of bleeding. To better address to patient needs, the strengths and weaknesses of each of these systems must be well-known, in order to have a personalized dialysis prescription for each patient.

Using Artificial Intelligence Resources in Dialysis and Kidney Transplant Patients: A Literature Review

BACKGROUND

The purpose of this review is to depict current research and impact of artificial intelligence/machine learning (AI/ML) algorithms on dialysis and kidney transplantation. Published studies were presented from two points of view: What medical aspects were covered? What AI/ML algorithms have been used?

METHODS

We searched four electronic databases or studies that used AI/ML in hemodialysis (HD), peritoneal dialysis (PD), and kidney transplantation (KT). Sixty-nine studies were split into three categories: AI/ML and HD, PD, and KT, respectively. We identified 43 trials in the first group, 8 in the second, and 18 in the third. Then, studies were classified according to the type of algorithm.

RESULTS

AI and HD trials covered: (a) dialysis service management, (b) dialysis procedure, (c) anemia management, (d) hormonal/dietary issues, and (e) arteriovenous fistula assessment. PD studies were divided into (a) peritoneal technique issues, (b) infections, and (c) cardiovascular event prediction. AI in transplantation studies were allocated into (a) management systems (ML used as pretransplant organ-matching tools), (b) predicting graft rejection, (c) tacrolimus therapy modulation, and (d) dietary issues.

CONCLUSIONS

Although guidelines are reluctant to recommend AI implementation in daily practice, there is plenty of evidence that AI/ML algorithms can predict better than nephrologists: volumes, Kt/V, and hypotension or cardiovascular events during dialysis. Altogether, these trials report a robust impact of AI/ML on quality of life and survival in G5D/T patients. In the coming years, one would probably witness the emergence of AI/ML devices that facilitate the management of dialysis patients, thus increasing the quality of life and survival.

A feedback system that combines monitoring of systolic blood pressure and relative blood volume in order to prevent hypotensive episodes during dialysis

Hypotensive Episodes (HEs) are one of the most common complications during dialysis. Occurrence of HEs can be reduced by applying physiological closed loop systems that monitor physiological parameter(s) and adjust dialysis related parameter(s). We developed a physiological closed loop control system (PCLCS) that monitors systolic blood pressure (sysBP) and relative blood volume (RBV) and calculates the net fluid removal (nfr) rate during dialysis. The performance of PCLCS was compared in the laboratory to a feedback system that monitors only RBV (BVFS). A laboratory test setup was developed to test the feedback systems. The test setup simulates nfr-rate and refilling of a patient's intravascular fluid. We studied the impact of the feedback systems PCLCS and BVFS on the number of HEs (sysBP < 90 mmHg), on the variance of sysBP and RBV, on pre to post sysBP and RBV and on the achievement of the nfr-volume. PCLCS allowed 80% less HEs than BVFS (p < 0.001). Variance of sysBP and RBV were reduced by 41.8% and by 52% (p < 0.001), respectively, when using PCLCS. There were no differences between pre to post sysBP nor between pre to post RBV when comparing PCLCS to BVFS. The nfr-volume was achieved by both feedback systems.

Changes in Red Blood Cell Size and Red Cell Fragmentation during Hemodialysis

Introduction

Intradialytic hypotension remains the commonest complication for outpatient hemodialysis. The majority of relative blood volume (RBV) monitoring techniques monitor changes in hematocrit. As hematocrit can potentially be affected by changes in red cell size and hemolysis we studied the change in red blood cell size (MCV) during dialysis and hemolysis.

Methods

MCV was prospectively measured in 176 stable regular adult hemodialysis outpatients (56% male, 27.8% diabetic, mean age 59.5 ±16.2 years) dialyzing against a range of dialysate sodiums (136–145 mmol/L), with cooled dialysate (35–36°C), containing 1 g/L glucose, with constant ultrafiltration profiles. Red cell fragmentation was studied in 41 of the cohort.

Results

Logistical regression analysis showed that the absolute change in MCV was related to the change in hematocrit (F=4.92, β=0.111, p=0.031), and inversely with red cell shrinkage associated with predialysis osmolality (F=5.06, β=0.83, p=0.029), and dialysate sodium (F=4.7, β=0.34, p=0.035). There was no significant increase in red cell fragments during the dialysis sessions.

Conclusions

Indirect assessment of RBV based upon the relative change in hematocrit, depends not only upon the change in plasma water and red blood cell numbers, but also upon MCV. Changes in MCV may theoretically lead to potential effects on RBV measurements.

Intra-dialytic blood oxygen saturation (SO2): association with dialysis hypotension (the SOGLIA Study)

BACKGROUND

Intradialytic hypotension (IDH) has a dramatic impact on the main outcomes of dialysis patients. Early warning of hemodynamic worsening during dialysis would enable preventive measures to be taken. Blood oxygen saturation (SO₂) is used for hemodynamic monitoring in the critical care setting and may provide useful information about IDH onset.

AIM

To evaluate whether short- and medium-term variations in the SO₂ signal (ST-SO_2var, MT-SO_2var,) during dialysis are a predictor of IDH.

METHODS

In this 3-month observational cohort study, 51 hypotension-prone chronic hemodialysis (HD) patients, with vascular access by arteriovenous fistula (AVF) or central venous catheter (CVC), were enrolled. Continuous non-invasive blood SO₂ was monitored (fc = 0.2 Hz) by an optical sensor on the arterial line of the extracorporeal circulation; blood pressure (every 30 min), symptoms and their time of appearance were noted. Predictive power of IDH was expressed by the area under curve (AUC) sensitivity and specificity based on intradialytic variations in SO₂.

RESULTS

A total of 1290 HD sessions were analyzed. Overall, off-line ST-SO_2var analysis proved able to correctly predict IDH in 67 % of the sessions where IDH occurred. The best predictive performance was found in the presence of highly arterialized AVF (SO₂ > 95 %) (75 % sensitivity; AUC 0.825; p < 0.05). On the contrary, in sessions with CVC, IDH prediction proved more efficient by MT-SO_2var (AUC 0.575; p = 0.01).

CONCLUSIONS

Intradialytic SO₂ variability could be a valid parameter to detect in advance the hemodynamic worsening that precedes IDH. Appropriate timely intervention could help prevent IDH onset.

Computer Aided Detection System for Prediction of the Malaise during Hemodialysis

Monitoring of dialysis sessions is crucial as different stress factors can yield suffering or critical situations. Specialized personnel is usually required for the administration of this medical treatment; nevertheless, subjects whose clinical status can be considered stable require different monitoring strategies when compared with subjects with critical clinical conditions. In this case domiciliary treatment or monitoring can substantially improve the quality of life of patients undergoing dialysis. In this work, we present a Computer Aided Detection (CAD) system for the telemonitoring of patients' clinical parameters. The CAD was mainly designed to predict the insurgence of critical events; it consisted of two Random Forest (RF) classifiers: the first one (RF₁) predicting the onset of any malaise one hour after the treatment start and the second one (RF₂) again two hours later. The developed system shows an accurate classification performance in terms of both sensitivity and specificity. The specificity in the identification of nonsymptomatic sessions and the sensitivity in the identification of symptomatic sessions for RF₂ are equal to 86.60% and 71.40%, respectively, thus suggesting the CAD as an effective tool to support expert nephrologists in telemonitoring the patients.

How to extract clinically useful information from large amount of dialysis related stored data

The basic storage infrastructure used to gather data from the technological evolution also in the healthcare field was leading to the storing into public or private repository of even higher quantities of data related to patients and their pathological evolution. Big data techniques are spreading also in medical research. By these techniques is possible extract information from complex heterogeneous sources, realizing longitudinal studies focused to correlate the patient status with biometric parameters. In our work we develop a common data infrastructure involving 4 clinical dialysis centers between Lombardy and Switzerland. The common platform has been build to store large amount of clinical data related to 716 dialysis session of 70 patient. The platform is made up by a combination of a MySQL(®) database (Dialysis Database) and a MATLAB-based mining library (Dialysis MATlib). A statistical analysis of these data has been performed on the data gathered. These analyses led to the development of two clinical indexes, representing an example of transformation of big data into clinical information.

Intradialytic hypotension and cardiac remodeling: a vicious cycle

Hemodynamic instability during hemodialysis is a common but often underestimated issue in the nephrologist practice. Intradialytic hypotension, namely, a decrease of systolic or mean blood pressure to a certain level, prohibits the safe and smooth achievement of ultrafiltration and solute removal goal in chronic dialysis patients. Studies have elucidated the potential mechanisms involved in the development of Intradialytic hypotension, including excessive ultrafiltration and loss of compensatory mechanisms for blood pressure maintenance. Cardiac remodeling could also be one important piece of the puzzle. In this review, we intend to discuss the role of cardiac remodeling, including left ventricular hypertrophy, in the development of Intradialytic hypotension. In addition, we will also provide evidence that a bidirectional relationship might exist between Intradialytic hypotension and left ventricular hypertrophy in chronic dialysis patients. A more complete understanding of the complex interactions in between could assist the readers in formulating potential solutions for the reduction of both phenomena.

Design and construct an optical device to determine relative blood volume in patients undergoing hemodialysis

Background:

Occurrence of hypotension during hemodialysis in nearly 20-30% of patients, shows is the necessity of continuous monitoring the patients' blood pressure during hemodialysis. Since directly and non-invasively continuous blood pressure monitoring, is not easy, finding a parameter related to blood pressure, for indirect monitoring is of great value. Related blood volume (RBV) is one of the parameters, related to blood pressure and have a good potential to reflect the patient’s hemodynamic condition.

Objectives:

The main objective of this study was to design and construct an optical device to determine the RBV in patients undergoing hemodialysis, during the process.

Materials and Methods:

After initial studies in order to select a proper sensor, using the ORCAD software, an analog circuit was designed. The implementation and modification of the circuit was done by the clinical tests, using expired blood. Afterwards, for calculation the RBV, controlling the display, data storage and sending it to the computer, an ATmega16 microcontroller was used. For programing the microcontroller, CodeVision software and then Altium Designer software were used for the circuit compression, in order to design the printed circuit board. Finally, all parts of the analog and digital circuit, AC to DC converter and the LCD were embedded in a box.

Results:

After finalization of the device and before testing it in a real situation, expired blood was used for final evaluation. The evaluation was done by changing the blood concentration, at the start point by adding water to it. In fact, the device can track the changes in blood concentration and display the RBV. After this evaluation, the device was tested in a clinical situation. The results showed there are no interactions between this device and the other devices used in the dialysis section and it can work properly in order to measure the RBV.

Conclusions:

Considering the hypotension and its consequences in a patient on hemodialysis, solving this problem seems necessary. One method for preventing this, is to use the blood pressure related parameters and one of these parameters is the RBV. In this study, in order to measure the RBV, a device was designed and evaluated by expired blood and also tested in a clinical situation. Results showed that the device could work properly in order to measure the RBV.

Mathematical biomarkers for the autonomic regulation of cardiovascular system

Heart rate and blood pressure are the most important vital signs in diagnosing disease. Both heart rate and blood pressure are characterized by a high degree of short term variability from moment to moment, medium term over the normal day and night as well as in the very long term over months to years. The study of new mathematical algorithms to evaluate the variability of these cardiovascular parameters has a high potential in the development of new methods for early detection of cardiovascular disease, to establish differential diagnosis with possible therapeutic consequences. The autonomic nervous system is a major player in the general adaptive reaction to stress and disease. The quantitative prediction of the autonomic interactions in multiple control loops pathways of cardiovascular system is directly applicable to clinical situations. Exploration of new multimodal analytical techniques for the variability of cardiovascular system may detect new approaches for deterministic parameter identification. A multimodal analysis of cardiovascular signals can be studied by evaluating their amplitudes, phases, time domain patterns, and sensitivity to imposed stimuli, i.e., drugs blocking the autonomic system. The causal effects, gains, and dynamic relationships may be studied through dynamical fuzzy logic models, such as the discrete-time model and discrete-event model. We expect an increase in accuracy of modeling and a better estimation of the heart rate and blood pressure time series, which could be of benefit for intelligent patient monitoring. We foresee that identifying quantitative mathematical biomarkers for autonomic nervous system will allow individual therapy adjustments to aim at the most favorable sympathetic-parasympathetic balance.

[Fluid overload and arterial hypertension in hemodialysis patients]

The water sodium overload is a factor of morbi-mortality and its treatment is one of the markers of adequacy of the hemodialysis treatment. Its first clinical assessment was improved by tools such as echocardiography and ultrasonography of the inferior vena cava, the per-dialytic curve of plasma volume, measuring BNP or proBNP and by impedancemetry. The combination of the evaluation of these parameters and of the clinical situation allows one to assess the extracellular overload, the state of the blood volume and the potential of plasma refilling. The latter is a key factor of the per-dialytic hemodynamic tolerance. It is itself a determining factor in weight can be achieved at the end of the session. Getting the "dry" weight can require modifications of the prescriptions of the hemodialysis sessions, a filling by albumin even a drugs support. Finally, the overload treatment is the central part of the treatment of arterial hypertension, which has to benefit however often from antihypertensive treatment the profit of which is demonstrated.

Noninvasive techniques for prevention of intradialytic hypotension

Episodes of hypotension during hemodialysis treatment constitutes an important clinical problem which has received considerable attention in recent years. Despite the fact that numerous approaches to reducing the frequency of intradialytic hypotension (IDH) have been proposed and evaluated, the problem has not yet found a definitive solution--an observation which, in particular, applies to episodes of acute, symptomatic hypotension. This overview covers recent advances in methodology for predicting and preventing IDH. Following a brief overview of well-established hypotension-related variables, including blood pressure, blood temperature, relative blood volume, and bioimpedance, special attention is given to electrocardiographic and photoplethysmographic (PPG) variables and their significance for IDH prediction. It is concluded that cardiovascular variables which reflect heart rate variability, heart rate turbulence, and baroreflex sensitivity are important to explore in feedback control hemodialysis systems so as to improve their performance. The analysis of hemodialysis-related changes in PPG pulse wave properties hold considerable promise for improving prediction.

Intra-dialytic hypotension and blood volume and blood temperature monitoring

Intra-dialytic hypotension (IDH) is a common problem affecting haemodialysis patients. Its aetiology is complex and influenced by multiple patient and dialysis factors. IDH occurs when the normal cardiovascular response cannot compensate for volume loss associated with ultrafiltration, and is exacerbated by a myriad of factors including intra-dialytic fluid gains, cardiovascular disease, antihypertensive medications and the physiological demands placed on patients by conventional haemodialysis. The use of blood volume monitoring and blood temperature monitoring technologies is advocated as a tool to predict and therefore prevent episodes of IDH. We review the clinical utility of these technologies and summarize the current evidence of their effect on reducing the incidence of IDH in haemodialysis population.

Bioimpedance-based identification of malnutrition using fuzzy logic

Protein-energy malnutrition reduces the quality of life, lengthens the time in hospital and dramatically increases mortality. Currently there is no simple and objective method available for assessing nutritional status and identifying malnutrition. The aim of this work is to develop a novel assistance system that supports the physician in the assessment of the nutritional status. Therefore, three subject groups were investigated: the first group consisted of 688 healthy subjects. Two additional groups consisted of 707 patients: 94 patients with primary diseases that are known to cause malnutrition, and 613 patients from a hospital admission screening. In all subjects bioimpedance spectroscopy measurements were performed, and the body composition was calculated. Additionally, in all patients the nutritional status was assessed by the subjective global assessment score. These data are used for the development and validation of the assistance system. The basic idea of the system is that nutritional status is reflected by body composition. Hence, features of the nutritional status, based on the body composition, are determined and compared with reference ranges, derived from healthy subjects' data. The differences are evaluated by a fuzzy logic system or a decision tree in order to identify malnourished patients. The novel assistance system allows the identification of malnourished patients, and it can be applied for screening and monitoring of the nutritional status of hospital patients.

Recent advances in the prevention and management of intradialytic hypotension

Intradialytic hypotension continues to play a significant role in the morbidity and in some cases the mortality associated with maintenance hemodialysis. Greater precision in the determination of dry weight using bioimpedance technology and biofeedback systems designed to prevent rapid fluctuations in blood volume have recently been shown to decrease the frequency of this complication. Pharmacologic strategies designed to maintain peripheral vascular resistance in patients with insufficient release of endogenous vasoconstrictors continue to be explored. The sudden development of intradialytic hypotension may respond to specific antagonists to hypotensive mediators.