Indirect measurement of lung density and air volume from electrical impedance tomography (EIT) data

Development of Machine Learning-Based Web System for Estimating Pleural Effusion Using Multi-Frequency Bioelectrical Impedance Analyses

BACKGROUND

Transthoracic impedance values have not been widely used to measure extravascular pulmonary water content due to accuracy and complexity concerns. Our aim was to develop a foundational model for a novel system aiming to non-invasively estimate the intrathoracic condition of heart failure patients.

METHODS

We employed multi-frequency bioelectrical impedance analysis to simultaneously measure multiple frequencies, collecting electrical, physical, and hematological data from 63 hospitalized heart failure patients and 82 healthy volunteers. Measurements were taken upon admission and after treatment, and longitudinal analysis was conducted.

RESULTS

Using a light gradient boosting machine, and a decision tree-based machine learning method, we developed an intrathoracic estimation model based on electrical measurements and clinical findings. Out of the 286 features collected, the model utilized 16 features. Notably, the developed model demonstrated high accuracy in discriminating patients with pleural effusion, achieving an area under the receiver characteristic curves (AUC) of 0.905 (95% CI: 0.870-0.940, p < 0.0001) in the cross-validation test. The accuracy significantly outperformed the conventional frequency-based method with an AUC of 0.740 (95% CI: 0.688-0.792, and p < 0.0001).

CONCLUSIONS

Our findings indicate the potential of machine learning and transthoracic impedance measurements for estimating pleural effusion. By incorporating noninvasive and easily obtainable clinical and laboratory findings, this approach offers an effective means of assessing intrathoracic conditions.

Evaluation of a Novel Contactless Electrical Impedance Device for Measuring Respiratory and Heart Rates: A Pilot Study

Although the respiratory rate is an important vital sign, it is rarely recorded in hospitals given the lack of convenient measurement devices. Posh Wellness Laboratory Inc. (Tokyo, Japan) developed a novel contactless card-type respiratory/heart rate monitoring device that measures electrical impedance variations on the human chest. This study was aimed to test and validate the accuracy of the proposed device compared with conventional medical monitors. To evaluate the card-type monitoring device, we compared the measurements from the device with those from the mechanical ventilators and electrocardiogram monitors. Patients who were hospitalized in the Emergency Department of Shimane University Hospital from April 5 to 30, 2021 were included in this study. A card-type sensor was attached to five patients who agreed to participate in this study. Four of the five patients were receiving mechanical ventilation. The respiratory rate error provided by the card-type sensor remained within 15% compared with the measurements of the conventional medical monitor. In contrast, the heart rate counts were largely different from the measurements. Thus, the proposed device can successfully measure the respiratory rate, whereas heart rate measurements require further improvement. Our small, lightweight, radiation-free, and contactless monitoring device can conveniently measure the respiratory rate of patients. With the improvement of measuring the heart rate, we would like to assess a larger number and a wider range of patients.

A Sensor for Spirometric Feedback in Ventilation Maneuvers during Cardiopulmonary Resuscitation Training

This work proposes adapting an existing sensor and embedding it on mannequins used in cardiopulmonary resuscitation (CPR) training to accurately measure the amount of air supplied to the lungs during ventilation. Mathematical modeling, calibration, and validation of the sensor along with metrology, statistical inference, and spirometry techniques were used as a base for aquiring scientific knowledge of the system. The system directly measures the variable of interest (air volume) and refers to spirometric techniques in the elaboration of its model. This improves the realism of the dummies during the CPR training, because it estimates, in real-time, not only the volume of air entering in the lungs but also the Forced Vital Capacity (FVC), Forced Expiratory Volume (FEVt) and Medium Forced Expiratory Flow (FEF_20–75%). The validation of the sensor achieved results that address the requirements for this application, that is, the error below 3.4% of full scale. During the spirometric tests, the system presented the measurement results of (305 ± 22, 450 ± 23, 603 ± 24, 751 ± 26, 922 ± 27, 1021 ± 30, 1182 ± 33, 1326 ± 36, 1476 ± 37, 1618 ± 45 and 1786 ± 56) × 10⁻⁶ m³ for reference values of (300, 450, 600, 750, 900, 1050, 1200, 1350, 1500, 1650 and 1800) × 10⁻⁶ m³, respectively. Therefore, considering the spirometry and pressure boundary conditions of the manikin lungs, the system achieves the objective of simulating valid spirometric data for debriefings, that is, there is an agreement between the measurement results when compared to the signal generated by a commercial spirometer (Koko brand). The main advantages that this work presents in relation to the sensors commonly used for this purpose are: (i) the reduced cost, which makes it possible, for the first time, to use a respiratory volume sensor in medical simulators or training dummies; (ii) the direct measurement of air entering the lung using a noninvasive method, which makes it possible to use spirometry parameters to characterize simulated human respiration during the CPR training; and (iii) the measurement of spirometric parameters (FVC, FEVt, and FEF_20–75%), in real-time, during the CPR training, to achieve optimal ventilation performance. Therefore, the system developed in this work addresses the minimum requirements for the practice of ventilation in the CPR maneuvers and has great potential in several future applications.

Global tidal variations, regional distribution of ventilation, and the regional onset of filling determined by electrical impedance tomography: reproducibility

The reproducibility of the regional distribution of ventilation and the timing of onset of regional filling as measured by electrical impedance tomography lacks evidence. This study investigated whether electrical impedance tomography measurements in healthy males were reproducible when electrodes were replaced between measurements. Part 1: Recordings of five volunteers lying supine were made using electrical impedance tomography and a pneumotachometer. Measurements were repeated at least three hours later. Skin marking ensured accurate replacement of electrodes. No stabilisation period was allowed. Part 2: Electrical impedance tomography recordings of ten volunteers; a 15 minute stabilisation period, extra skin markings, and time-averaging were incorporated to improve the reproducibility. Reproducibility was determined using the Bland-Altman method. To judge the transferability of the limits of agreement, a Pearson correlation was used for electrical impedance tomography tidal variation and tidal volume. Tidal variation was judged to be reproducible due to the significant correlation between tidal variation and tidal volume (r² = 0.93). The ventilation distribution was not reproducible. A stabilisation period, extra skin markings and time-averaging did not improve the outcome. The timing of regional onset of filling was reproducible and could prove clinically valuable. The reproducibility of the tidal variation indicates that non-reproducibility of the ventilation distribution was probably a biological difference and not measurement error. Other causes of variability such as electrode placement variability or lack of stabilisation when accounted for did not improve the reproducibility of the ventilation distribution.

Coupling of EIT with computational lung modeling for predicting patient-specific ventilatory responses

Providing optimal personalized mechanical ventilation for patients with acute or chronic respiratory failure is still a challenge within a clinical setting for each case anew. In this article, we integrate electrical impedance tomography (EIT) monitoring into a powerful patient-specific computational lung model to create an approach for personalizing protective ventilatory treatment. The underlying computational lung model is based on a single computed tomography scan and able to predict global airflow quantities, as well as local tissue aeration and strains for any ventilation maneuver. For validation, a novel “virtual EIT” module is added to our computational lung model, allowing to simulate EIT images based on the patient's thorax geometry and the results of our numerically predicted tissue aeration. Clinically measured EIT images are not used to calibrate the computational model. Thus they provide an independent method to validate the computational predictions at high temporal resolution. The performance of this coupling approach has been tested in an example patient with acute respiratory distress syndrome. The method shows good agreement between computationally predicted and clinically measured airflow data and EIT images. These results imply that the proposed framework can be used for numerical prediction of patient-specific responses to certain therapeutic measures before applying them to an actual patient. In the long run, definition of patient-specific optimal ventilation protocols might be assisted by computational modeling.

NEW & NOTEWORTHY In this work, we present a patient-specific computational lung model that is able to predict global and local ventilatory quantities for a given patient and any selected ventilation protocol. For the first time, such a predictive lung model is equipped with a virtual electrical impedance tomography module allowing real-time validation of the computed results with the patient measurements. First promising results obtained in an acute respiratory distress syndrome patient show the potential of this approach for personalized computationally guided optimization of mechanical ventilation in future.