Detecting the symptoms of Parkinson's disease with non-standard video.

BACKGROUND: Neurodegenerative diseases, such as Parkinson's disease (PD), necessitate frequent clinical visits and monitoring to identify changes in motor symptoms and provide appropriate care. By applying machine learning techniques to video data, automated video analysis has emerged as a promising approach to track and analyze motor symptoms, which could facilitate more timely intervention. However, existing solutions often rely on specialized equipment and recording procedures, which limits their usability in unstructured settings like the home. In this study, we developed a method to detect PD symptoms from unstructured videos of clinical assessments, without the need for specialized equipment or recording procedures. METHODS: Twenty-eight individuals with Parkinson's disease completed a video-recorded motor examination that included the finger-to-nose and hand pronation-supination tasks. Clinical staff provided ground truth scores for the level of Parkinsonian symptoms present. For each video, we used a pre-existing model called PIXIE to measure the location of several joints on the person's body and quantify how they were moving. Features derived from the joint angles and trajectories, designed to be robust to recording angle, were then used to train two types of machine-learning classifiers (random forests and support vector machines) to detect the presence of PD symptoms. RESULTS: The support vector machine trained on the finger-to-nose task had an F1 score of 0.93 while the random forest trained on the same task yielded an F1 score of 0.85. The support vector machine and random forest trained on the hand pronation-supination task had F1 scores of 0.20 and 0.33, respectively. CONCLUSION: These results demonstrate the feasibility of developing video analysis tools to track motor symptoms across variable perspectives. These tools do not work equally well for all tasks, however. This technology has the potential to overcome barriers to access for many individuals with degenerative neurological diseases like PD, providing them with a more convenient and timely method to monitor symptom progression, without requiring a structured video recording procedure. Ultimately, more frequent and objective home assessments of motor function could enable more precise telehealth optimization of interventions to improve clinical outcomes inside and outside of the clinic.

Video-Based Pose Estimation for Gait Analysis in Stroke Survivors during Clinical Assessments: A Proof-of-Concept Study.

Recent advancements in deep learning have produced significant progress in markerless human pose estimation, making it possible to estimate human kinematics from single camera videos without the need for reflective markers and specialized labs equipped with motion capture systems. Such algorithms have the potential to enable the quantification of clinical metrics from videos recorded with a handheld camera. Here we used DeepLabCut, an open-source framework for markerless pose estimation, to fine-tune a deep network to track 5 body keypoints (hip, knee, ankle, heel, and toe) in 82 below-waist videos of 8 patients with stroke performing overground walking during clinical assessments. We trained the pose estimation model by labeling the keypoints in 2 frames per video and then trained a convolutional neural network to estimate 5 clinically relevant gait parameters (cadence, double support time, swing time, stance time, and walking speed) from the trajectory of these keypoints. These results were then compared to those obtained from a clinical system for gait analysis (GAITRite®, CIR Systems). Absolute accuracy (mean error) and precision (standard deviation of error) for swing, stance, and double support time were within 0.04 ± 0.11 s; Pearson's correlation with the reference system was moderate for swing times (r = 0.4-0.66), but stronger for stance and double support time (r = 0.93-0.95). Cadence mean error was -0.25 steps/min ± 3.9 steps/min (r = 0.97), while walking speed mean error was -0.02 ± 0.11 m/s (r = 0.92). These preliminary results suggest that single camera videos and pose estimation models based on deep networks could be used to quantify clinically relevant gait metrics in individuals poststroke, even while using assistive devices in uncontrolled environments. Such development opens the door to applications for gait analysis both inside and outside of clinical settings, without the need of sophisticated equipment.

Combining Accelerometer and GPS Features to Evaluate Community Mobility in Knee Ankle Foot Orthoses (KAFO) Users.

Orthotic and assistive devices such as knee ankle foot orthoses (KAFO), come in a variety of forms and fits, with several levels of available features that could help users perform daily activities more naturally. However, objective data on the actual use of these devices outside of the research lab is usually not obtained. Such data could enhance traditional lab-based outcome measures and inform clinical decision-making when prescribing new orthotic and assistive technology. Here, we link data from a GPS unit and an accelerometer mounted on the orthotic device to quantify its usage in the community and examine the correlations with clinical metrics. We collected data from 14 individuals over a period of 2 months as they used their personal KAFO first, and then a novel research KAFO; for each device we quantified number of steps, cadence, time spent at community locations and time wearing the KAFO at those locations. Sensor-derived metrics showed that mobility patterns differed widely between participants (mean steps: 591.3, SD =704.2). The novel KAFO generally enabled participants to walk faster during clinical tests ( ∆6 Minute-Walk-Test=71.5m, p=0.006). However, some participants wore the novel device less often despite improved performance on these clinical measures, leading to poor correlation between changes in clinical outcome measures and changes in community mobility ( ∆6 Minute-Walk-Test - ∆ Community Steps: r=0.09, p=0.76). Our results suggest that some traditional clinical outcome measures may not be associated with the actual wear time of an assistive device in the community, and obtaining personalized data from real-world use through wearable technology is valuable.

Rapid Screening of Physiological Changes Associated With COVID-19 Using Soft-Wearables and Structured Activities: A Pilot Study.

OBJECTIVE: Controlling the spread of the COVID-19 pandemic largely depends on scaling up the testing infrastructure for identifying infected individuals. Consumer-grade wearables may present a solution to detect the presence of infections in the population, but the current paradigm requires collecting physiological data continuously and for long periods of time on each individual, which poses limitations in the context of rapid screening. Technology: Here, we propose a novel paradigm based on recording the physiological responses elicited by a short (~2 minutes) sequence of activities (i.e. "snapshot"), to detect symptoms associated with COVID-19. We employed a novel body-conforming soft wearable sensor placed on the suprasternal notch to capture data on physical activity, cardio-respiratory function, and cough sounds. RESULTS: We performed a pilot study in a cohort of individuals (n=14) who tested positive for COVID-19 and detected altered heart rate, respiration rate and heart rate variability, relative to a group of healthy individuals (n=14) with no known exposure. Logistic regression classifiers were trained on individual and combined sets of physiological features (heartbeat and respiration dynamics, walking cadence, and cough frequency spectrum) at discriminating COVID-positive participants from the healthy group. Combining features yielded an AUC of 0.94 (95% CI=[0.92, 0.96]) using a leave-one-subject-out cross validation scheme. Conclusions and Clinical Impact: These results, although preliminary, suggest that a sensor-based snapshot paradigm may be a promising approach for non-invasive and repeatable testing to alert individuals that need further screening.

Role of data measurement characteristics in the accurate detection of Parkinson's disease symptoms using wearable sensors.

BACKGROUND: Parkinson's disease (PD) is a progressive neurological disease, with characteristic motor symptoms such as tremor and bradykinesia. There is a growing interest to continuously monitor these and other symptoms through body-worn sensor technology. However, limited battery life and memory capacity hinder the potential for continuous, long-term monitoring with these devices. There is little information available on the relative value of adding sensors, increasing sampling rate, or computing complex signal features, all of which may improve accuracy of symptom detection at the expense of computational resources. Here we build on a previous study to investigate the relationship between data measurement characteristics and accuracy when using wearable sensor data to classify tremor and bradykinesia in patients with PD. METHODS: Thirteen individuals with PD wore a flexible, skin-mounted sensor (collecting tri-axial accelerometer and gyroscope data) and a commercial smart watch (collecting tri-axial accelerometer data) on their predominantly affected hand. The participants performed a series of standardized motor tasks, during which a clinician scored the severity of tremor and bradykinesia in that limb. Machine learning models were trained on scored data to classify tremor and bradykinesia. Model performance was compared when using different types of sensors (accelerometer and/or gyroscope), different data sampling rates (up to 62.5 Hz), and different categories of pre-engineered features (up to 148 features). Performance was also compared between the flexible sensor and smart watch for each analysis. RESULTS: First, there was no effect of device type for classifying tremor symptoms (p > 0.34), but bradykinesia models incorporating gyroscope data performed slightly better (up to 0.05 AUROC) than other models (p = 0.01). Second, model performance decreased with sampling frequency (p < 0.001) for tremor, but not bradykinesia (p > 0.47). Finally, model performance for both symptoms was maintained after substantially reducing the feature set. CONCLUSIONS: Our findings demonstrate the ability to simplify measurement characteristics from body-worn sensors while maintaining performance in PD symptom detection. Understanding the trade-off between model performance and data resolution is crucial to design efficient, accurate wearable sensing systems. This approach may improve the feasibility of long-term, continuous, and real-time monitoring of PD symptoms by reducing computational burden on wearable devices.

Association of sleep with neurobehavioral impairments during inpatient rehabilitation after traumatic brain injury.

BACKGROUND: Sleep disturbance is a common sequela after traumatic brain injury (TBI). Many of the impairments following TBI may be exacerbated by impaired sleep-wake cycle regulation. OBJECTIVES: To investigate the relationship between total sleep time (TST), measured by wrist actigraphy and observational sleep logs, and neurobehavioral impairments during inpatient rehabilitation after TBI. METHODS: Twenty-five subjects undergoing inpatient rehabilitation for traumatic brain injury were included. TST was measured using wrist actigraphy and observational sleep logs. Neurobehavioral impairments were assessed using the Neurobehavioral Rating Scale-Revised (NRS-R), a multidimensional clinician-based assessment. RESULTS: Of 25 subjects enrolled, 23 subjects completed the study. A significant negative correlation was found between total NRS-R and TST calculated by observational sleep logs (r = -0.28, p = 0.007). The association between total NRS-R and TST, as calculated by actigraphy, was not significantly correlated (R = -0.01, p = 0.921). CONCLUSIONS: Sleep disturbance during inpatient rehabilitation is associated with neurobehavioral impairments after TBI. TST measured by actigraphy may be limited by sleep detection algorithms that have not been validated in certain patient populations. Considerations should be made regarding the feasibility of using wearable sensors in patients with cognitive and behavioral impairments. Challenges regarding actigraphy for sleep monitoring in the brain injury population are discussed.

Wearable sensors for Parkinson's disease: which data are worth collecting for training symptom detection models.

Machine learning algorithms that use data streams captured from soft wearable sensors have the potential to automatically detect PD symptoms and inform clinicians about the progression of disease. However, these algorithms must be trained with annotated data from clinical experts who can recognize symptoms, and collecting such data are costly. Understanding how many sensors and how much labeled data are required is key to successfully deploying these models outside of the clinic. Here we recorded movement data using 6 flexible wearable sensors in 20 individuals with PD over the course of multiple clinical assessments conducted on 1 day and repeated 2 weeks later. Participants performed 13 common tasks, such as walking or typing, and a clinician rated the severity of symptoms (bradykinesia and tremor). We then trained convolutional neural networks and statistical ensembles to detect whether a segment of movement showed signs of bradykinesia or tremor based on data from tasks performed by other individuals. Our results show that a single wearable sensor on the back of the hand is sufficient for detecting bradykinesia and tremor in the upper extremities, whereas using sensors on both sides does not improve performance. Increasing the amount of training data by adding other individuals can lead to improved performance, but repeating assessments with the same individuals-even at different medication states-does not substantially improve detection across days. Our results suggest that PD symptoms can be detected during a variety of activities and are best modeled by a dataset incorporating many individuals.

Addendum of: Fall Detection in Individuals With Lower Limb Amputations Using Mobile Phones: Machine Learning Enhances Robustness for Real-World Applications.

[This corrects the article DOI: 10.2196/mhealth.8201.].

Fall Detection in Individuals With Lower Limb Amputations Using Mobile Phones: Machine Learning Enhances Robustness for Real-World Applications.

BACKGROUND: Automatically detecting falls with mobile phones provides an opportunity for rapid response to injuries and better knowledge of what precipitated the fall and its consequences. This is beneficial for populations that are prone to falling, such as people with lower limb amputations. Prior studies have focused on fall detection in able-bodied individuals using data from a laboratory setting. Such approaches may provide a limited ability to detect falls in amputees and in real-world scenarios. OBJECTIVE: The aim was to develop a classifier that uses data from able-bodied individuals to detect falls in individuals with a lower limb amputation, while they freely carry the mobile phone in different locations and during free-living. METHODS: We obtained 861 simulated indoor and outdoor falls from 10 young control (non-amputee) individuals and 6 individuals with a lower limb amputation. In addition, we recorded a broad database of activities of daily living, including data from three participants' free-living routines. Sensor readings (accelerometer and gyroscope) from a mobile phone were recorded as participants freely carried it in three common locations-on the waist, in a pocket, and in the hand. A set of 40 features were computed from the sensors data and four classifiers were trained and combined through stacking to detect falls. We compared the performance of two population-specific models, trained and tested on either able-bodied or amputee participants, with that of a model trained on able-bodied participants and tested on amputees. A simple threshold-based classifier was used to benchmark our machine-learning classifier. RESULTS: The accuracy of fall detection in amputees for a model trained on control individuals (sensitivity: mean 0.989, 1.96*standard error of the mean [SEM] 0.017; specificity: mean 0.968, SEM 0.025) was not statistically different (P=.69) from that of a model trained on the amputee population (sensitivity: mean 0.984, SEM 0.016; specificity: mean 0.965, SEM 0.022). Detection of falls in control individuals yielded similar results (sensitivity: mean 0.979, SEM 0.022; specificity: mean 0.991, SEM 0.012). A mean 2.2 (SD 1.7) false alarms per day were obtained when evaluating the model (vs mean 122.1, SD 166.1 based on thresholds) on data recorded as participants carried the phone during their daily routine for two or more days. Machine-learning classifiers outperformed the threshold-based one (P<.001). CONCLUSIONS: A mobile phone-based fall detection model can use data from non-amputee individuals to detect falls in individuals walking with a prosthesis. We successfully detected falls when the mobile phone was carried across multiple locations and without a predetermined orientation. Furthermore, the number of false alarms yielded by the model over a longer period of time was reasonably low. This moves the application of mobile phone-based fall detection systems closer to a real-world use case scenario.

Activity Recognition in Individuals Walking With Assistive Devices: The Benefits of Device-Specific Models.

BACKGROUND: Wearable sensors gather data that machine-learning models can convert into an identification of physical activities, a clinically relevant outcome measure. However, when individuals with disabilities upgrade to a new walking assistive device, their gait patterns can change, which could affect the accuracy of activity recognition. OBJECTIVE: The objective of this study was to assess whether we need to train an activity recognition model with labeled data from activities performed with the new assistive device, rather than data from the original device or from healthy individuals. METHODS: Data were collected from 11 healthy controls as well as from 11 age-matched individuals with disabilities who used a standard stance control knee-ankle-foot orthosis (KAFO), and then a computer-controlled adaptive KAFO (Ottobock C-Brace). All subjects performed a structured set of functional activities while wearing an accelerometer on their waist, and random forest classifiers were used as activity classification models. We examined both global models, which are trained on other subjects (healthy or disabled individuals), and personal models, which are trained and tested on the same subject. RESULTS: Median accuracies of global and personal models trained with data from the new KAFO were significantly higher (61% and 76%, respectively) than those of models that use data from the original KAFO (55% and 66%, respectively) (Wilcoxon signed-rank test, P=.006 and P=.01). These models also massively outperformed a global model trained on healthy subjects, which only achieved a median accuracy of 53%. Device-specific models conferred a major advantage for activity recognition. CONCLUSIONS: Our results suggest that when patients use a new assistive device, labeled data from activities performed with the specific device are needed for maximal precision activity recognition. Personal device-specific models yield the highest accuracy in such scenarios, whereas models trained on healthy individuals perform poorly and should not be used in patient populations.

Sensor Fusion to Infer Locations of Standing and Reaching Within the Home in Incomplete Spinal Cord Injury.

OBJECTIVE: The objective of rehabilitation after spinal cord injury is to enable successful function in everyday life and independence at home. Clinical tests can assess whether patients are able to execute functional movements but are limited in assessing such information at home. A prototype system is developed that detects stand-to-reach activities, a movement with important functional implications, at multiple locations within a mock kitchen. DESIGN: Ten individuals with incomplete spinal cord injuries performed a sequence of standing and reaching tasks. The system monitored their movements by combining two sources of information: a triaxial accelerometer, placed on the subject's thigh, detected sitting or standing, and a network of radio frequency tags, wirelessly connected to a wrist-worn device, detected reaching at three locations. A threshold-based algorithm detected execution of the combined tasks and accuracy was measured by the number of correctly identified events. RESULTS: The system was shown to have an average accuracy of 98% for inferring when individuals performed stand-to-reach activities at each tag location within the same room. CONCLUSIONS: The combination of accelerometry and tags yielded accurate assessments of functional stand-to-reach activities within a home environment. Optimization of this technology could simplify patient compliance and allow clinicians to assess functional home activities.

The need to approximate the use-case in clinical machine learning.

The availability of smartphone and wearable sensor technology is leading to a rapid accumulation of human subject data, and machine learning is emerging as a technique to map those data into clinical predictions. As machine learning algorithms are increasingly used to support clinical decision making, it is vital to reliably quantify their prediction accuracy. Cross-validation (CV) is the standard approach where the accuracy of such algorithms is evaluated on part of the data the algorithm has not seen during training. However, for this procedure to be meaningful, the relationship between the training and the validation set should mimic the relationship between the training set and the dataset expected for the clinical use. Here we compared two popular CV methods: record-wise and subject-wise. While the subject-wise method mirrors the clinically relevant use-case scenario of diagnosis in newly recruited subjects, the record-wise strategy has no such interpretation. Using both a publicly available dataset and a simulation, we found that record-wise CV often massively overestimates the prediction accuracy of the algorithms. We also conducted a systematic review of the relevant literature, and found that this overly optimistic method was used by almost half of the retrieved studies that used accelerometers, wearable sensors, or smartphones to predict clinical outcomes. As we move towards an era of machine learning-based diagnosis and treatment, using proper methods to evaluate their accuracy is crucial, as inaccurate results can mislead both clinicians and data scientists.

Using and understanding cross-validation strategies. Perspectives on Saeb et al.

This three-part review takes a detailed look at the complexities of cross-validation, fostered by the peer review of Saeb et al.'s paper entitled "The need to approximate the use-case in clinical machine learning." It contains perspectives by reviewers and by the original authors that touch upon cross-validation: the suitability of different strategies and their interpretation.

Accelerometry-enabled measurement of walking performance with a robotic exoskeleton: a pilot study.

BACKGROUND: Clinical scores for evaluating walking skills with lower limb exoskeletons are often based on a single variable, such as distance walked or speed, even in cases where a host of features are measured. We investigated how to combine multiple features such that the resulting score has high discriminatory power, in particular with few patients. A new score is introduced that allows quantifying the walking ability of patients with spinal cord injury when using a powered exoskeleton. METHODS: Four spinal cord injury patients were trained to walk over ground with the ReWalk™ exoskeleton. Body accelerations during use of the device were recorded by a wearable accelerometer and 4 features to evaluate walking skills were computed. The new score is the Gaussian naïve Bayes surprise, which evaluates patients relative to the features' distribution measured in 7 expert users of the ReWalk™. We compared our score based on all the features with a standard outcome measure, which is based on number of steps only. RESULTS: All 4 patients improved over the course of training, as their scores trended towards the expert users' scores. The combined score (Gaussian naïve surprise) was considerably more discriminative than the one using only walked distance (steps). At the end of training, 3 out of 4 patients were significantly different from the experts, according to the combined score (p < .001, Wilcoxon Signed-Rank Test). In contrast, all but one patient were scored as experts when number of steps was the only feature. CONCLUSION: Integrating multiple features could provide a more robust metric to measure patients' skills while they learn to walk with a robotic exoskeleton. Testing this approach with other features and more subjects remains as future work.

Activity recognition in patients with lower limb impairments: do we need training data from each patient?

Machine learning allows detecting specific physical activities using data from wearable sensors. Such a quantification of patient mobility over time promises to accurately inform clinical decisions for physical rehabilitation. There are two strategies of setting up the machine learning problem: detect one patient's activities using data from the same patient (personal model) or detect their activities using data from other patients (global model), and we currently do not know if personal models are necessary. Here we consider the problem of detecting physical activities from a waist-worn accelerometer in patients who use a knee-ankle-foot orthosis (KAFO) to walk. We show that while a model based on healthy subjects has low accuracy, the global model performs as well as the personal model. This is encouraging because it suggests that condition-specific activity recognition algorithms are sufficient and that no data from individual patients is necessary.

Robust active binocular vision through intrinsically motivated learning.

The efficient coding hypothesis posits that sensory systems of animals strive to encode sensory signals efficiently by taking into account the redundancies in them. This principle has been very successful in explaining response properties of visual sensory neurons as adaptations to the statistics of natural images. Recently, we have begun to extend the efficient coding hypothesis to active perception through a form of intrinsically motivated learning: a sensory model learns an efficient code for the sensory signals while a reinforcement learner generates movements of the sense organs to improve the encoding of the signals. To this end, it receives an intrinsically generated reinforcement signal indicating how well the sensory model encodes the data. This approach has been tested in the context of binocular vison, leading to the autonomous development of disparity tuning and vergence control. Here we systematically investigate the robustness of the new approach in the context of a binocular vision system implemented on a robot. Robustness is an important aspect that reflects the ability of the system to deal with unmodeled disturbances or events, such as insults to the system that displace the stereo cameras. To demonstrate the robustness of our method and its ability to self-calibrate, we introduce various perturbations and test if and how the system recovers from them. We find that (1) the system can fully recover from a perturbation that can be compensated through the system's motor degrees of freedom, (2) performance degrades gracefully if the system cannot use its motor degrees of freedom to compensate for the perturbation, and (3) recovery from a perturbation is improved if both the sensory encoding and the behavior policy can adapt to the perturbation. Overall, this work demonstrates that our intrinsically motivated learning approach for efficient coding in active perception gives rise to a self-calibrating perceptual system of high robustness.

Incremental learning control of the DLR-HIT-Hand II during interaction tasks.

In this paper a bio-inspired control architecture for a robotic hand is presented. It relies on the same mechanisms of learning inverse internal models studied in humans. The control is capable of developing an internal representation of the hand interacting with the environment and updating it by means of the interaction forces that arise during contact. The learning paradigm exploits LWPR networks, which allow efficient incremental online learning through the use of spatially localized linear regression models. Additionally this paradigm limits negative interference when learning multiple tasks. The architecture is validated on a simulated finger of the DLR-HIT-Hand II performing closing movements in presence of two different viscous force fields, perturbing its motion.

An internal model for acquisition and retention of motor learning during arm reaching.

Humans have the ability to learn novel motor tasks while manipulating the environment. Several models of motor learning have been proposed in the literature, but few of them address the problem of retention and interference of motor memory. The modular selection and identification for control (MOSAIC) model, originally proposed by Wolpert and Kawato, is one of the most relevant contributions; it suggests a possible strategy on how the human motor control system learns and adapts to novel environments. MOSAIC employs the concept of forward and inverse models. The same group later proposed the hidden Markov model (HMM) MOSAIC, which affords learning multiple tasks. The significant drawback of this second approach is that the HMM must be trained with a complete data set that includes all contexts. Since the number of contexts or modules is fixed from the onset, this approach does not afford incremental learning of new tasks. In this letter, we present an alternative architecture to overcome this problem, based on a nonparametric regression algorithm, named locally weighted projection regression (LWPR). This network structure develops according to the contexts allowing incremental training. Of notice, interaction force is used to disambiguate among different contexts. We demonstrate the capability of this alternative architecture with a simulated 2 degree-of-freedom representation of the human arm that learns to interact with three distinct objects, reproducing the same test paradigm of the HMM MOSAIC. After learning the dynamics of the three objects, the LWPR network successfully learns to compensate for a novel velocity-dependent force field. Equally important, it retains previously acquired knowledge on the interaction with the three objects. Thus, this architecture allows both incremental learning of new tasks and retention of previously acquired knowledge, a feature of human motor learning and memory.

Dispensing an enzyme-conjugated solution into an ELISA plate by adapting ink-jet printers.

The rapid and precise delivery of small volumes of bio-fluids (from picoliters to nanoliters) is a key feature of modern bioanalytical assays. Commercial ink-jet printers are low-cost systems which enable the dispensing of tiny droplets at a rate which may exceed 10(4) Hz per nozzle. Currently, the main ejection technologies are piezoelectric and bubble-jet. We adapted two commercial printers, respectively a piezoelectric and a bubble-jet one, for the deposition of immunoglobulins into an ELISA plate. The objective was to perform a comparative evaluation of the two classes of ink-jet technologies in terms of required hardware modifications and possible damage on the dispensed molecules. The hardware of the two printers was modified to dispense an enzyme conjugate solution, containing polyclonal rabbit anti-human IgG labelled with HRP in 7 wells of an ELISA plate. Moreover, the ELISA assay was used to assess the functional activity of the biomolecules after ejection. ELISA is a common and well-assessed technique to detect the presence of particular antigens or antibodies in a sample. We employed an ELISA diagnostic kit for the qualitative screening of anti-ENA antibodies to verify the ability of the dispensed immunoglobulins to bind the primary antibodies in the wells. Experimental tests showed that the dispensing of immunoglobulins using the piezoelectric printer does not cause any detectable difference on the outcome of the ELISA test if compared to manual dispensing using micropipettes. On the contrary, the thermal printhead was not able to reliably dispense the bio-fluid, which may mean that a surfactant is required to modify the wetting properties of the liquid.