Jie Hou

• Laeng, Bruno; Zambrana, Imac; Hou, Jie; Martinsen, Ørjan Grøttem & Holm, Kristiane (2026). Machine learning classification of emotional videos via autonomic arousal based on pupil diameters, electrodermal activity and heart rate. International Journal of Psychophysiology. ISSN 0167-8760. p. 113455–113455. doi: 10.1016/j.ijpsycho.2026.113455. Full text in Research Archive Show summary

  We evaluated the ability of Electrodermal Activity (EDA), Heart Rate Variability (HRV), and Pupil Diameter (PD) to detect emotional states via machine learning to determine whether simple arousal metrics can distinguish complex emotional labels: Amusement, Awe, Disgust, Nurturant Love, and Sadness. In a laboratory session, participants (N = 47) selected the most appropriate emotional label out of five immediately after viewing 10 short videos. Classifications relied solely on the three psychophysiological measures and by applying three machine learning models: Support Vector Machine (SVM/SVC), Random Forest (RF), and Extreme Gradient Boosting (XGBoost). Features included baseline-corrected average pupil diameter (PD) and electrodermal activity (EDA) peaks. Noisy heart rate variability (HRV) data were excluded to prevent signal-to-noise ratio degradation. PD reflects sympathetic noradrenergic activation and reciprocal coupling within the Autonomic Nervous System (ANS), providing a stable state estimate. In the core ‘affect circumplex’, SVC achieved peak accuracy, and PD achieved the highest feature importance score across architectures. Results revealed a many-to-one mapping where discrete valence states produced indistinguishable physiological signatures: high-arousal states (Amusement, Sadness) and low-arousal states (Awe, Disgust) were conflated. Notably, Awe triggered PD constriction, mirroring Disgust's profile. Tactical removal of the cardiovascular features significantly improved RF and XGBoost performance. These findings suggest that simple arousal metrics, and specifically pupillometry, can classify affect without using complex time-series analysis. Non-invasive sensors may facilitate the development of sensitive computerized systems calibrated to objective physiological data.

• Hou, Jie; Ali, Abdulkadir Hassen & Martinsen, Ørjan Grøttem (2025). Using neuromorphic computing in prediction of GABA concentration – a pilot study. Journal of Electrical Bioimpedance. 16(1), p. 146–149. doi: 10.2478/joeb-2025-0019. Full text in Research Archive Show summary

  Abstract Neuromorphic computing has the potential to facilitate detection of GABA concentration levels in the brain, and offers energy-efficient, real-time machine learning processing possibilities. To study whether neuromorphic computing can be used for GABA concentration detection, dielectric relaxation spectroscopy was used to acquire permittivity data of different concentrations of GABA solution. Thereafter, two different machine learning models were compared (Feedforward neural network (FFNN) and convolutional neural network (CNN)) for accuracy in prediction of GABA concentration from dielectric properties. The CNN model was then converted to spiking Neural Networks (SNNs), which showed promising results for energy efficiency and real-time processing capabilities. The system incorporates Tkinter, a Python interface to the Tcl/Tk GUI toolkit for seamless data transfer between the neuromorphic chip and the measurement system, ensuring flexibility and scalability in a user-friendly system.

• Færden, Ida Høy; Bliksøen, Marte; Liavåg, Olav Magnus Ivar; Hou, Jie; Majeed, Waleed Mohammed & Torp, May-Kristin [Show all 14 contributors for this article] (2025). Organ monitoring and identification of ischemic liver injury during ex situ porcine liver machine perfusion—A randomized controlled study. Liver transplantation. ISSN 1527-6465. 32(2), p. 153–165. doi: 10.1097/lvt.0000000000000752. Full text in Research Archive Show summary

  The global shortage of donor livers has led to an increased reliance on extended-criteria donor livers, including those from donors after circulatory death. Utilizing marginal livers necessitates distinguishing between injured but recoverable livers and those that are irreversibly damaged. Bile duct injury, prevalent in these grafts, poses significant risks, as ischemic injuries can be challenging to assess and may lead to recipient bile duct strictures and ultimately graft failure. This study aimed to identify objective measurements that enhance existing viability criteria and facilitate the diagnosis of ischemic injuries. We employed ex situ machine perfusion (MP) to monitor porcine livers subjected to standardized biliary injury (BileINJ), standardized global liver injury (GlobalINJ), and no injury (CTRL, total n=23). Using microdialysis catheters and pCO 2 sensors, we analyzed liver tissue, hilar plate, and bile duct, in addition to standard viability assessments, collecting data on 77 variables. Principal component analysis indicated distinct clustering of the GlobalINJ and partially the BileINJ group away from the CTRL group during MP. Feature extraction models highlighted microdialysate lactate, lactate/pyruvate ratio, glycerol, tissue pCO 2 , and blood gas hematocrit as critical indicators for classifying liver state. Histopathological evaluations confirmed group-specific liver and bile injuries. We identified unique metabolic patterns that differentiate ischemically injured from non-injured porcine livers and were able to distinguish liver and biliary injury. Real-time monitoring of livers using microdialysis and tissue pCO 2 during MP is feasible and clinically available. Measurements taken in the hilar plate show the ability to identify bile duct damage while not introducing measurement devices to the vulnerable bile duct itself. Our findings enable more objective and timely selection of transplantable livers and warrant further investigation in clinical studies.

• Bari, Dindar S.; Rammoo, Mohammed Noor S.; Youssif, Ardawan A.; Najman, Hoger M.; Aldosky, Haval Y. Yacoob & Tronstad, Christian [Show all 8 contributors for this article] (2024). Electrodermal Activity for Quantitative Assessment of Dental Anxiety. Journal of Sensor and Actuator Networks. 13(6). doi: 10.3390/jsan13060088. Full text in Research Archive Show summary

  In spite of the development in technology and the recent innovations in dentistry, dental anxiety remains a common issue, and accurately assessing it is challenging due to reliance on patients’ self-reports, which are often biased. Hence, this study was undertaken to determine whether dental anxiety can be quantified objectively using the EDA parameters. EDA (skin conductance (SC), skin susceptance (SS), and skin potential (SP)) parameters and heart rate (HR) were recorded from 40 participants during two different sessions (baseline and anxiety). In addition, the Modified Dental Anxiety Scale (MDAS) scale was also used to record the level of anxiety. The physiological data from EDA and HR were compared with the subjective self-reports of anxiety provided on the MDAS to show whether higher EDA and HR readings correspond to higher scores on the MDAS. To elicit dental anxiety, participants were exposed to several film clips associated with dental treatment. EDA signals were compared between the two sessions for all clips and all EDA scores. SC and HR significantly (p < 0.05) increased during the anxiety session compared to the baseline session. The number of fluctuations per minute in the SC, SS, and SP notably increased during the anxiety session. The MDAS results revealed that the participants had dental anxiety when they were exposed to video clips of dental procedures. The study results imply that EDA parameters could be used as a useful tool to monitor dental anxiety, in particular in young children and non-verbal patients or those with intellectual disabilities, which may aid the dentist in the successful management of dental anxiety during treatment. The moment-to-moment EDA data on a patient’s anxiety levels throughout particular dental operations provides a clearer picture of anxiety variations than pre- or post-appointment surveys alone, in addition to offering unbiased tracking of dental anxiety levels over self-reports. This study seeks to encourage further research into the most effective EDA parameters for improving the management of dental anxiety.

• Abie, Sisay Mebre; Román, Alejandro Ortega de & Hou, Jie (2024). Effect of body orientation and joint movement on local bioimpedance measurements. Journal of Electrical Bioimpedance. 15(1), p. 137–144. doi: 10.2478/joeb-2024-0016. Full text in Research Archive Show summary

  The purpose of this pilot study was to determine if body orientation, skin treatment, joint angle, and shoulder arch movements affect localized bioimpedance spectroscopy (BIS) measurement. Nowadays, there are various wearable and portable impedance measurement tools in different shapes and sizes. Therefore, the body position and orientation of the subject during measurement may be of great importance for the comparability of the results. Ultrasound machine was used to measure the thickness of the skin layers and then bioimpedance measurements were performed for ten young men (age=23 ± 5) at room temperature (22°C) for different body orientations, skin treatments, joint angle, and shoulder arch movements. The results were analyzed using statistical methods and graphical presentation using Python and MatLab. Our observations indicate that there is a significant difference between standing straight up, supine and sitting positions. The results show that there is a significant difference between the two skin treatments (alcohol vs tape stripping). Moreover, joint angle and shoulder arch movements also have an impact on the impedance data. Therefore, to be able to control these factors can potentially improve the quality and comparability of the measured impedance data.

• Hou, Jie; Strand-Amundsen, Runar James & Martinsen, Ørjan Grøttem (2023). Detection of physiological concentrations of GABA using dielectric spectroscopy - A pilot study. Journal of Electrical Bioimpedance. 14(1), p. 47–52. doi: 10.2478/joeb-2023-0006. Full text in Research Archive Show summary

  Abstract Gamma-aminobutyric acid (GABA) is a major inhibitory neurotransmitter that is present at a relatively low level throughout the normal adult human brain. Abnormal GABA levels are found in people with neurodegenerative disorders such as Parkinson’s disease, epilepsy, schizophrenia, depression, and others. Being able to measure the GABA concentration would be beneficial for patient groups with fluctuating GABA levels for better diagnosis and treatment. In this study, we explore the feasibility of using dielectric relaxation spectroscopy for the detection of GABA concentrations within a physiological range, with the perspective of miniaturization and use during implantation. Utilizing machine learning techniques, we were able to differentiate GABA concentrations down to 5 μm. This work investigates a novel use of dielectric relaxation spectroscopy, to assess if physiological GABA concentrations can be detected through permittivity measurements.

• Dapsance, Florian; Hou, Jie; Dufour, Damien; Boccara, Charlotte; Briand, Nolwenn & Martinsen, Ørjan Grøttem (2023). Adipose Tissue Characterization With Electrical Impedance Spectroscopy and Machine Learning. IEEE Sensors Letters. 7(10). doi: 10.1109/LSENS.2023.3317921. Full text in Research Archive Show summary

  Biological tissues have variable passive electrical properties depending on their cellular constitution. Electrical impedance spectroscopy (EIS) is commonly used to monitor cell and tissue characteristics. By measuring the impedance of a sample at various frequencies, it is possible to collect information regarding cell size and shape, cell membrane properties, or cytoplasm conductivity. From the perspective of longitudinal structural monitoring, bioimpedance measurements outrank traditional tissue analysis methods, such as fixation and slicing, owing to their nondestructive nature. Machine learning can be used to automatically process the impedance data and make real-time classifications of tissue types. Here, we present preliminary results on ex-vivo mouse adipose tissue measurements using EIS and further data processing and classification using machine learning models.

• Hou, Jie; Liavåg, Olav Magnus Ivar; Færden, Ida Høy; Martinsen, Ørjan Grøttem; Tønnessen, Tor Inge & Line, Pål Dag [Show all 10 contributors for this article] (2022). Utilization of dielectric properties for assessment of liver ischemia-reperfusion injury in vivo and during machine perfusion. Scientific Reports. 12(1). doi: 10.1038/s41598-022-14817-3. Full text in Research Archive Show summary

  Abstract There is a shortage of donor livers and patients consequently die on waiting lists worldwide. Livers are discarded if they are clinically judged to have a high risk of non-function following transplantation. With the aim of extending the pool of available donor livers, we assessed the condition of porcine livers by monitoring the microwave dielectric properties. A total of 21 livers were divided into three groups: control with no injury (CON), biliary injury by hepatic artery occlusion (AHEP), and overall hepatic injury by static cold storage (SCS). All were monitored for four hours in vivo, followed by ex vivo plurithermic machine perfusion (PMP). Permittivity data was modeled with a two-pole Cole–Cole equation, and dielectric properties from one-hour intervals were analyzed during in vivo and normothermic machine perfusion (NMP). A clear increasing trend in the conductivity was observed in vivo in the AHEP livers compared to the control livers. After four hours of NMP, separations in the conductivity were observed between the three groups. Our results indicate that dielectric relaxation spectroscopy (DRS) can be used to detect and differentiate liver injuries, opening for a standardized and reliable point of evaluation for livers prior to transplantation.

• Hou, Jie; Strand-Amundsen, Runar James; Tronstad, Christian; Tønnessen, Tor Inge; Høgetveit, Jan Olav & Martinsen, Ørjan Grøttem (2022). Small intestinal viability assessment using dielectric relaxation spectroscopy and deep learning. Scientific Reports. 12(1). doi: 10.1038/s41598-022-07140-4. Full text in Research Archive Show summary

  Abstract Intestinal ischemia is a serious condition where the surgeon often has to make important but difficult decisions regarding resections and resection margins. Previous studies have shown that 3 h (hours) of warm full ischemia of the small bowel followed by reperfusion appears to be the upper limit for viability in the porcine mesenteric ischemia model. However, the critical transition between 3 to 4 h of ischemic injury can be nearly impossible to distinguish intraoperatively based on standard clinical methods. In this study, permittivity data from porcine intestine was used to analyze the characteristics of various degrees of ischemia/reperfusion injury. Our results show that dielectric relaxation spectroscopy can be used to assess intestinal viability. The dielectric constant and conductivity showed clear differences between healthy, ischemic and reperfused intestinal segments. This indicates that dielectric parameters can be used to characterize different intestinal conditions. In addition, machine learning models were employed to classify viable and non-viable segments based on frequency dependent dielectric properties of the intestinal tissue, providing a method for fast and accurate intraoperative surgical decision-making. An average classification accuracy of 98.7% was obtained using only permittivity data measured during ischemia, and 96.2% was obtained with data measured during reperfusion. The proposed approach allows the surgeon to get accurate evaluation from the trained machine learning model by performing one single measurement on an intestinal segment where the viability state is questionable.

• Hou, Jie; Ness, Siri Schøne; Tschudi, Jon; O'Farrell, Marion; Veddegjerde, Rune & Martinsen, Ørjan Grøttem [Show all 8 contributors for this article] (2022). Assessment of Intestinal Ischemia–Reperfusion Injury Using Diffuse Reflectance VIS-NIR Spectroscopy and Histology. Sensors. 22(23). doi: 10.3390/s22239111. Full text in Research Archive Show summary

  A porcine model was used to investigate the feasibility of using VIS-NIR spectroscopy to differentiate between degrees of ischemia–reperfusion injury in the small intestine. Ten pigs were used in this study and four segments were created in the small intestine of each pig: (1) control, (2) full arterial and venous mesenteric occlusion for 8 h, (3) arterial and venous mesenteric occlusion for 2 h followed by reperfusion for 6 h, and (4) arterial and venous mesenteric occlusion for 4 h followed by reperfusion for 4 h. Two models were built using partial least square discriminant analysis. The first model was able to differentiate between the control, ischemic, and reperfused intestinal segments with an average accuracy of 99.2% with 10-fold cross-validation, and the second model was able to discriminate between the viable versus non-viable intestinal segments with an average accuracy of 96.0% using 10-fold cross-validation. Moreover, histopathology was used to investigate the borderline between viable and non-viable intestinal segments. The VIS-NIR spectroscopy method together with a PLS-DA model showed promising results and appears to be well-suited as a potentially real-time intraoperative method for assessing intestinal ischemia–reperfusion injury, due to its easy-to-use and non-invasive nature.

• Bordvik, David Andreas; Hou, Jie; Noori, Farzan Majeed; Uddin, Md Zia & Tørresen, Jim (2022). Monitoring In-Home Emergency Situation and Preserve Privacy using Multi-modal Sensing and Deep Learning, 2022 International Conference on Electronics, Information, and Communication (ICEIC 2022). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781665409353. doi: 10.1109/ICEIC54506.2022.9748829. Full text in Research Archive Show summary

  Videos and images are commonly used in home monitoring systems. However, detecting emergencies in-home while preserving privacy is a challenging task concerning Human Activity Recognition (HAR). In recent years, HAR combined with deep learning has drawn much attention from the general public. Besides that, relying entirely on a single sensor modal-ity is not promising. In this paper, depth images and radar presence data were used to investigate if such sensor data can tackle the challenge of a system's ability to detect abnormal and normal situations while preserving privacy. The recurrence plots and wavelet transformations were used to make a two-dimensional representation of the presence radar data. Moreover, we fused data from both sensors using data-level, feature-level, and decision-level fusions. The decision-level fusion showed its superiority over the other two techniques. For the decision-level fusion, a combination of the depth images and presence data recurrence plots trained first on convolutional neural networks (CNN). The output was fed into support vector machines, which yielded the best accuracy of 99.98%.

• Hou, Jie; Strand-Amundsen, Runar James; Hødnebø, Stina; Tønnessen, Tor Inge & Høgetveit, Jan Olav (2021). Assessing ischemic injury in human intestine ex vivo with electrical impedance spectroscopy. Journal of Electrical Bioimpedance. 12(1), p. 82–88. doi: 10.2478/joeb-2021-0011. Full text in Research Archive Show summary

  Abstract Electrical impedance spectroscopy is a well-established tool for monitoring changes in the electrical properties of tissue. Most tissue and organ types have been investigated in various studies. As for the small intestine, there are several published studies conducted on pig and rat models. This study investigates the changes in passive electrical properties of the complete wall of the human intestine non-invasively during ischemia. We aim to use the passive electrical properties to assess intestinal viability. The bioimpedance measurements were performed using a two-electrode set-up with a Solartron 1260 Impedance/gain-phase analyser. The small intestinal samples were resected from patients who underwent pancreaticoduodenectomy. Impedance measurements were conducted following resection by placing the electrodes on the surface of the intestine. A voltage was applied across the intestinal sample and the measured electrical impedance was obtained in the ZPlot software. Impedance data were further fitted into a Cole model to obtain the Cole parameters. The P y value was calculated from the extracted Cole parameters and used to assess the cell membrane integrity, thus evaluate the intestinal viability. Eight small intestinal segments from different patients were used in this study and impedance measurements were performed once an hour for a ten-hour period. One hour after resection, the impedance decreased, then increased the next two hours, before decreasing until the end of the experiment. For all the intestinal segments, the P y values first increased and reached a plateau which lasted for 1 - 2 hours, before it decreased irreversibly. The time interval where P y value reached the maximum is consistent with reported viable/non-viable limits from histological analysis.

• Singstad, Bjørn Jostein; Azulay, Naomi; Bjurstedt, Andreas; Bjørndal, Simen Chill; Drageseth, Magnus Flø & Austdal, Lars Peter Engeset [Show all 21 contributors for this article] (2021). Estimation of heart rate variability from finger photoplethysmography during rest, mild exercise and mild mental stress. Journal of Electrical Bioimpedance. 12(1), p. 89–102. doi: 10.2478/joeb-2021-0012. Full text in Research Archive Show summary

  Abstract Due to the possibilities in miniaturization and wearability, photoplethysmography (PPG) has recently gained a large interest not only for heart rate measurement, but also for estimating heart rate variability, which is derived from ECG by convention. The agreement between PPG and ECG-based HRV has been assessed in several studies, but the feasibility of PPG-based HRV estimation is still largely unknown for many conditions. In this study, we assess the feasibility of HRV estimation based on finger PPG during rest, mild physical exercise and mild mental stress. In addition, we compare different variants of signal processing methods including selection of fiducial point and outlier correction. Based on five minutes synchronous recordings of PPG and ECG from 15 healthy participants during each of these three conditions, the PPG-based HRV estimation was assessed for the SDNN and RMSSD parameters, calculated based on two different fiducial points (foot point and maximum slope), with and without outlier correction. The results show that HRV estimation based on finger PPG is feasible during rest and mild mental stress, but can give large errors during mild physical exercise. A good estimation is very dependent on outlier correction and fiducial point selection, and SDNN seems to be a more robust parameter compared to RMSSD for PPG-based HRV estimation.

• Hou, Jie; Strand-Amundsen, Runar James; Tronstad, Christian; Høgetveit, Jan Olav; Martinsen, Ørjan Grøttem & Tønnessen, Tor Inge (2021). Automatic prediction of ischemia-reperfusion injury of small intestine using convolutional neural networks: A pilot study. Sensors. 21(19). doi: 10.3390/s21196691. Full text in Research Archive Show summary

  Acute intestinal ischemia is a life-threatening condition. The current gold standard, with evaluation based on visual and tactile sensation, has low specificity. In this study, we explore the feasibility of using machine learning models on images of the intestine, to assess small intestinal viability. A digital microscope was used to acquire images of the jejunum in 10 pigs. Ischemic segments were created by local clamping (approximately 30 cm in width) of small arteries and veins in the mesentery and reperfusion was initiated by releasing the clamps. A series of images were acquired once an hour on the surface of each of the segments. The convolutional neural network (CNN) has previously been used to classify medical images, while knowledge is lacking whether CNNs have potential to classify ischemia-reperfusion injury on the small intestine. We compared how different deep learning models perform for this task. Moreover, the Shapley additive explanations (SHAP) method within explainable artificial intelligence (AI) was used to identify features that the model utilizes as important in classification of different ischemic injury degrees. To be able to assess to what extent we can trust our deep learning model decisions is critical in a clinical setting. A probabilistic model Bayesian CNN was implemented to estimate the model uncertainty which provides a confidence measure of our model decisions.

• Abie, Sisay Mebre; Martinsen, Ørjan Grøttem; Egelandsdal, Bjørg; Hou, Jie; Bjerke, Frøydis & Mason, Alex [Show all 7 contributors for this article] (2021). Feasibility of using electrical impedance spectroscopy for assessing biological cell damage during freezing and thawing. Sensors. 21(12). doi: 10.3390/s21124129. Full text in Research Archive Show summary

  This study was performed to test bioimpedance as a tool to detect the effect of different thawing methods on meat quality to aid in the eventual creation of an electric impedance-based food quality monitoring system. The electric impedance was measured for fresh pork, thawed pork, and during quick and slow thawing. A clear difference was observed between fresh and thawed samples for both impedance parameters. Impedance was different between the fresh and the frozen-thawed samples, but there were no impedance differences between frozen-thawed samples and the ones that were frozen-thawed and then stored at +3 °C for an additional 16 h after thawing. The phase angle was also different between fresh and the frozen-thawed samples. At high frequency, there were small, but clear phase angle differences between frozen-thawed samples and the samples that were frozen-thawed and subsequently stored for more than 16 h at +3 °C. Furthermore, the deep learning model LSTM-RNN (long short-term memory recurrent neural network) was found to be a promising way to classify the different methods of thawing.

• Hou, Jie; Abie, Sisay Mebre; Strand-Amundsen, Runar James; Galperin, Yuri; Bergli, Joakim & Schuelke, Christin [Show all 8 contributors for this article] (2021). Temperature dependence of the microwave dielectric properties of γ -aminobutyric acid. Scientific Reports. 11(1). doi: 10.1038/s41598-021-97178-7. Full text in Research Archive Show summary

  The GABA molecule is the major inhibitory neurotransmitter in the mammalian central nervous system. Through binding to post-synaptic neurons, GABA reduces the neuronal excitability by hyperpolarization. Correct binding between the GABA molecules and its receptors relies on molecular recognition. Earlier studies suggest that recognition is determined by the geometries of the molecule and its receptor. We employed dielectric relaxation spectroscopy (DRS) to study the conformation and dielectric properties of the GABA molecule under physiologically relevant laboratory conditions. The dielectric properties of GABA investigated have given us new insights about the GABA molecule, such as how they interact with each other and with water molecules at different temperatures (22°C and 37.5°C). Higher temperature leads to lower viscosity, thus lower relaxation time. The change in the GABA relaxation time due to concentration change is more associated with the solution viscosity than with the GABA dipole moment. A resonance behavior was observed with high GABA concentrations at physiological temperature, where there might be a phase transition at a certain temperature for a given GABA concentration that leads to a sudden change of the dielectric properties.

• Cunha, Andrè Bastos da; Hou, Jie & Schuelke, Christin (2019). Machine learning for stem cell differentiation and proliferation classification on electrical impedance spectroscopy. Journal of Electrical Bioimpedance. 10(1), p. 124–132. doi: 10.2478/joeb-2019-0018. Full text in Research Archive Show summary

  Electrical impedance spectroscopy (EIS) measurements on cells is a proven method to assess stem cell proliferation and differentiation. Cell regenerative medicine (CRM) is an emerging field where the need to develop and deploy stem cell assessment techniques is paramount as experimental treatments reach pre-clinical and clinical stages. However, EIS measurements on cells is a method requiring extensive post-processing and analysis. As a contribution to address this concern, we developed three machine learning models for three different stem cell lines able to classify the measured data as proliferation or differentiation laying the stone for future studies on using machine learning to profile EIS measurements on stem cells spectra.

View all works in NVA

• Hou, Jie; Nesaragi, Naimahmed & Tronstad, Christian (2024). Electrical bioimpedance in the era of artificial intelligence. Journal of Electrical Bioimpedance. 15(1), p. 1–3. doi: 10.2478/joeb-2024-0001. Full text in Research Archive

View all works in NVA