Multicenter prospective blinded melanoma detection study with a handheld elastic scattering spectroscopy device

Background

The elastic scattering spectroscopy (ESS) device (DermaSensor Inc., Miami, FL) is a noninvasive, painless, adjunctive tool for skin cancer detection.

Objectives

To investigate the performance of the ESS device in the detection of melanoma.

Methods

A prospective, investigator-blinded, multicenter study was conducted at 8 United States (US) and 2 Australian sites. All eligible skin lesions were clinically concerning for melanoma, examined with the ESS device, subsequently biopsied according to dermatologists' standard of care, and evaluated with histopathology. A total of 311 participants with 440 lesions were enrolled, including 44 melanomas (63.6% in situ and 36.4% invasive) and 44 severely dysplastic nevi.

Results

The observed sensitivity of the ESS device for melanoma detection was 95.5% (95% CI, 84.5% to 98.8%, 42 of 44 melanomas), and the observed specificity was 32.5% (95% CI, 27.2% to 38.3%). The positive and negative predictive values were 16.0% and 98.1%, respectively.

Limitations

The device was tested in a high-risk population with lesions selected for biopsy based on clinical and dermoscopic assessments of board-certified dermatologists. Most enrolled lesions were pigmented.

Conclusion

The ESS device's high sensitivity and NPV for the detection of melanoma suggest the device may be a useful adjunctive, point-of-care tool for melanoma detection.

Increased Flow Limitation During Sleep Is Associated With Increased Psychomotor Vigilance Task Lapses in Individuals With Suspected OSA

BACKGROUND

Impaired daytime vigilance is an important consequence of OSA, but several studies have reported no association between objective measurements of vigilance and the apnea-hypopnea index (AHI). Notably, the AHI does not quantify the degree of flow limitation, that is, the extent to which ventilation fails to meet intended ventilation (ventilatory drive).

RESEARCH QUESTION

Is flow limitation during sleep associated with daytime vigilance in OSA?

STUDY DESIGN AND METHODS

Nine hundred ninety-eight participants with suspected OSA completed a 10-min psychomotor vigilance task (PVT) before same-night in-laboratory polysomnography. Flow limitation frequency (percent of flow-limited breaths) during sleep was quantified using airflow shapes (eg, fluttering and scooping) from nasal pressure airflow. Multivariable regression assessed the association between flow limitation frequency and the number of lapses (response times > 500 ms, primary outcome), adjusting for age, sex, BMI, total sleep time, depression, and smoking status.

RESULTS

Increased flow limitation frequency was associated with decreased vigilance: a 1-SD (35.3%) increase was associated with 2.1 additional PVT lapses (95% CI, 0.7-3.7; P = .003). This magnitude was similar to that for age, where a 1-SD increase (13.5 years) was associated with 1.9 additional lapses. Results were similar after adjusting for AHI, hypoxemia severity, and arousal severity. The AHI was not associated with PVT lapses (P = .20). In secondary exploratory analysis, flow limitation frequency was associated with mean response speed (P = .012), median response time (P = .029), fastest 10% response time (P = .041), slowest 10% response time (P = .018), and slowest 10% response speed (P = .005).

INTERPRETATION

Increased flow limitation during sleep was associated with decreased daytime vigilance in individuals with suspected OSA, independent of the AHI. Flow limitation may complement standard clinical metrics in identifying individuals whose vigilance impairment most likely is explained by OSA.

Automated analysis of the total choline resonance peak in breast proton magnetic resonance spectroscopy

The aim of the current study was to compare the performance of fully automated software with human expert interpretation of single-voxel proton magnetic resonance spectroscopy (1H-MRS) spectra in the assessment of breast lesions. Breast magnetic resonance imaging (MRI) (including contrast-enhanced T1-weighted, T2-weighted, and diffusion-weighted imaging) and 1H-MRS images of 74 consecutive patients were acquired on a 3-T positron emission tomography-MRI scanner then automatically imported into and analyzed by SpecTec-ULR 1.1 software (LifeTec Solutions GmbH). All ensuing 117 spectra were additionally independently analyzed and interpreted by two blinded radiologists. Histopathology of at least 24 months of imaging follow-up served as the reference standard. Nonparametric Spearman's correlation coefficients for all measured parameters (signal-to-noise ratio [SNR] and integral of total choline [tCho]), Passing and Bablok regression, and receiver operating characteristic analysis, were calculated to assess test diagnostic performance, as well as to compare automated with manual reading. Based on 117 spectra of 74 patients, the area under the curve for tCho SNR and integrals ranged from 0.768 to 0.814 and from 0.721 to 0.784 to distinguish benign from malignant tissue, respectively. Neither method displayed significant differences between measurements (automated vs. human expert readers, p > 0.05), in line with the results from the univariate Spearman's rank correlation coefficients, as well as the Passing and Bablok regression analysis. It was concluded that this pilot study demonstrates that 1H-MRS data from breast MRI can be automatically exported and interpreted by SpecTec-ULR 1.1 software. The diagnostic performance of this software was not inferior to human expert readers.

Fully Automated Agatston Score Calculation From Electrocardiography-Gated Cardiac Computed Tomography Using Deep Learning and Multi-Organ Segmentation: A Validation Study

To evaluate deep learning-based calcium segmentation and quantification on ECG-gated cardiac CT scans compared with manual evaluation. Automated calcium quantification was performed using a neural network based on mask regions with convolutional neural networks (R-CNNs) for multi-organ segmentation. Manual evaluation of calcium was carried out using proprietary software. This is a retrospective study of archived data. This study used 40 patients to train the segmentation model and 110 patients were used for the validation of the algorithm. The Pearson correlation coefficient between the reference actual and the computed predictive scores shows high level of correlation (0.84; P < .001) and high limits of agreement (±1.96 SD; -2000, 2000) in Bland-Altman plot analysis. The proposed method correctly classifies the risk group in 75.2% and classifies the subjects in the same group. In total, 81% of the predictive scores lie in the same categories and only seven patients out of 110 were more than one category off. For the presence/absence of coronary artery calcifications, the deep learning model achieved a sensitivity of 90% and a specificity of 94%. Fully automated model shows good correlation compared with reference standards. Automating process reduces evaluation time and optimizes clinical calcium scoring without additional resources.

RAPID aneurysm accurately measures aneurysm size on CT angiography compared to three-dimensional digital subtraction angiography

BACKGROUND

Cerebral aneurysms are often identified and characterized on non-invasive CT Angiography (CTA) images, but digital subtraction angiography (DSA) is the gold standard for aneurysm evaluation.

OBJECTIVE

We compared cerebral aneurysm size measurements as measured from CTA processed by a semi-automated artificial intelligence software program (RAPID Aneurysm) and three-dimensional rotational DSA (3D-DSA).

METHODS

We performed a retrospective cohort study of consecutive patients with a cerebral aneurysm who underwent CTA and DSA with 3D reformations. CTA images were processed by RAPID Aneurysm to determine aneurysm height, width, and neck width. The reference standard was aneurysm measurements on 3D-DSA as measured by two neurointerventionalists. Both readers were blinded to RAPID Aneurysm measurements. Correlation and bias between these measurements were determined.

RESULTS

Results from 50 patients with 50 aneurysms were compared. 32 patients (64%) were female. Median age was 65 (IQR: 56.25-71.75). 37 patients (74%) presented with ruptured aneurysms. The aneurysms represented a range of aneurysm sizes (1.9-33.3 mm; IQR 3.6-7.2 mm). RAPID Aneurysm size measurements showed excellent correlation and low bias (correlation, mean difference) when compared to the reference standard for aneurysm height (0.98, -0.9 mm), width (0.98, 0.1 mm), and neck width (0.94, 1.1 mm). The inter-reader comparison between the two neurointerventionalists was similarly excellent for aneurysm height (0.97, -0.4 mm), width (0.98, -0.2 mm), and neck width (0.89, 0.8 mm).

CONCLUSION

RAPID Aneurysm measurement of cerebral aneurysm height, width, and neck width on CTA is strongly correlated to expert neurointerventionalist measurements on 3D-DSA.

Deep Learning Artificial Intelligence Tool for Automated Radiographic Determination of Posterior Tibial Slope in Patients With ACL Injury

Background

An increased posterior tibial slope (PTS) corresponds with an increased risk of graft failure after anterior cruciate ligament (ACL) reconstruction (ACLR). Validated methods of manual PTS measurements are subject to potential interobserver variability and can be inefficient on large datasets.

Purpose/Hypothesis

To develop a deep learning artificial intelligence technique for automated PTS measurement from standard lateral knee radiographs. It was hypothesized that this deep learning tool would be able to measure the PTS on a high volume of radiographs expeditiously and that these measurements would be similar to previously validated manual measurements.

Study Design

Cohort study (diagnosis); Level of evidence, 2.

Methods

A deep learning U-Net model was developed on a cohort of 300 postoperative short-leg lateral radiographs from patients who underwent ACLR to segment the tibial shaft, tibial joint surface, and tibial tuberosity. The model was trained via a random split after an 80 to 20 train-validation scheme. Masks for training images were manually segmented, and the model was trained for 400 epochs. An image processing pipeline was then deployed to annotate and measure the PTS using the predicted segmentation masks. Finally, the performance of this combined pipeline was compared with human measurements performed by 2 study personnel using a previously validated manual technique for measuring the PTS on short-leg lateral radiographs on an independent test set consisting of both pre- and postoperative images.

Results

The U-Net semantic segmentation model achieved a mean Dice similarity coefficient of 0.885 on the validation cohort. The mean difference between the human-made and computer-vision measurements was 1.92° (σ = 2.81° [P = .24]). Extreme disagreements between the human and machine measurements, as defined by ≥5° differences, occurred <5% of the time. The model was incorporated into a web-based digital application front-end for demonstration purposes, which can measure a single uploaded image in Portable Network Graphics format in a mean time of 5 seconds.

Conclusion

We developed an efficient and reliable deep learning computer vision algorithm to automate the PTS measurement on short-leg lateral knee radiographs. This tool, which demonstrated good agreement with human annotations, represents an effective clinical adjunct for measuring the PTS as part of the preoperative assessment of patients with ACL injuries.

Biofeedback and Digitalized Motivational Interviewing to Increase Daily Physical Activity: Series of Factorial N-of-1 Randomized Controlled Trials Piloting the Precious App

BACKGROUND

Insufficient physical activity is a public health concern. New technologies may improve physical activity levels and enable the identification of its predictors with high accuracy. The Precious smartphone app was developed to investigate the effect of specific modular intervention elements on physical activity and examine theory-based predictors within individuals.

OBJECTIVE

This study pilot-tested a fully automated factorial N-of-1 randomized controlled trial (RCT) with the Precious app and examined whether digitalized motivational interviewing (dMI) and heart rate variability-based biofeedback features increased objectively recorded steps. The secondary aim was to assess whether daily self-efficacy and motivation predicted within-person variability in daily steps.

METHODS

In total, 15 adults recruited from newspaper advertisements participated in a 40-day factorial N-of-1 RCT. They installed 2 study apps on their phones: one to receive intervention elements and one to collect ecological momentary assessment (EMA) data on self-efficacy, motivation, perceived barriers, pain, and illness. Steps were tracked using Xiaomi Mi Band activity bracelets. The factorial design included seven 2-day biofeedback interventions with a Firstbeat Bodyguard 2 (Firstbeat Technologies Ltd) heart rate variability sensor, seven 2-day dMI interventions, a wash-out day after each intervention, and 11 control days. EMA questions were sent twice per day. The effects of self-efficacy, motivation, and the interventions on subsequent steps were analyzed using within-person dynamic regression models and aggregated data using longitudinal multilevel modeling (level 1: daily observations; level 2: participants). The analyses were adjusted for covariates (ie, within- and between-person perceived barriers, pain or illness, time trends, and recurring events).

RESULTS

All participants completed the study, and adherence to activity bracelets and EMA measurements was high. The implementation of the factorial design was successful, with the dMI features used, on average, 5.1 (SD 1.0) times of the 7 available interventions. Biofeedback interventions were used, on average, 5.7 (SD 1.4) times out of 7, although 3 participants used this feature a day later than suggested and 1 did not use it at all. Neither within- nor between-person analyses revealed significant intervention effects on step counts. Self-efficacy predicted steps in 27% (4/15) of the participants. Motivation predicted steps in 20% (3/15) of the participants. Aggregated data showed significant group-level effects of day-level self-efficacy (B=0.462; P<.001), motivation (B=0.390; P<.001), and pain or illness (B=-1524; P<.001) on daily steps.

CONCLUSIONS

The automated factorial N-of-1 trial with the Precious app was mostly feasible and acceptable, especially the automated delivery of the dMI components, whereas self-conducted biofeedback measurements were more difficult to time correctly. The findings suggest that changes in self-efficacy and motivation may have same-day effects on physical activity, but the effects vary across individuals. This study provides recommendations based on the lessons learned on the implementation of factorial N-of-1 RCTs.

Assessing the Accuracy and Comprehensiveness of ChatGPT in Offering Clinical Guidance for Atopic Dermatitis and Acne Vulgaris

Collapse

Automated Image Analysis for Characterization of Circulating Tumor Cells and Clusters Sorted by Magnetic Levitation

Magnetic levitation-based sorting technologies have revolutionized the detection and isolation of rare cells, including circulating tumor cells (CTCs) and circulating tumor cell clusters (CTCCs). Manual counting and quantification of these cells are prone to time-consuming processes, human error, and inter-observer variability, particularly challenging when heterogeneous cell types in 3D clusters are present. To overcome these challenges, we developed "Fastcount," an in-house MATLAB-based algorithm for precise, automated quantification and phenotypic characterization of CTCs and CTCCs, in both 2D and 3D. Fastcount is 120 times faster than manual counting and produces reliable results with a ±7.3% deviation compared to a trained laboratory technician. By analyzing 400 GB of fluorescence imaging data, we showed that Fastcount outperforms manual counting and commercial software when cells are aggregated in 3D or staining artifacts are present, delivering more accurate results. We further employed Fastcount for automated analysis of 3D image stacks obtained from CTCCs isolated from colorectal adenocarcinoma and renal cell carcinoma blood samples. Interestingly, we observed a highly heterogeneous spatial cellular composition within CTCCs, even among clusters from the same patient. Overall, Fastcount can be employed for various applications with lab-chip devices, such as CTC detection, CTCC analysis in 3D and cell detection in biosensors.

Creating rare epilepsy cohorts using keyword search in electronic health records

OBJECTIVE

Administrative codes to identify people with rare epilepsies in electronic health records are limited. The current study evaluated the use of keyword search as an alternative method for rare epilepsy cohort creation using electronic health records data.

METHODS

Data included clinical notes from encounters with International Classification of Diseases, Ninth Revision (ICD-9) codes for seizures, epilepsy, and/or convulsions during 2010-2014, across six health care systems in New York City. We identified cases with rare epilepsies by searching clinical notes for keywords associated with 33 rare epilepsies. We validated cases via manual chart review. We compared the performance of keyword search to manual chart review using positive predictive value (PPV), sensitivity, and F-score. We selected an initial combination of keywords using the highest F-scores.

RESULTS

Data included clinical notes from 77 924 cases with ICD-9 codes for seizures, epilepsy, and/or convulsions. The all-keyword search method identified 6095 candidates, and manual chart review confirmed that 2068 (34%) had a rare epilepsy. The initial combination method identified 1862 cases with a rare epilepsy, and this method performed as follows: PPV median = .64 (interquartile range [IQR] = .50-.81, range = .20-1.00), sensitivity median = .93 (IQR = .76-1.00, range = .10-1.00), and F-score median = .71 (IQR = .63-.85, range = .18-1.00). Using this method, we identified four cohorts of rare epilepsies with over 100 individuals, including infantile spasms, Lennox-Gastaut syndrome, Rett syndrome, and tuberous sclerosis complex. We identified over 50 individuals with two rare epilepsies that do not have specific ICD-10 codes for cohort creation (epilepsy with myoclonic atonic seizures, Sturge-Weber syndrome).

SIGNIFICANCE

Keyword search is an effective method for cohort creation. These findings can improve identification and surveillance of individuals with rare epilepsies and promote their referral to specialty clinics, clinical research, and support groups.

An optimized HEK293T cell expansion protocol using a hollow-fiber bioreactor system

Viral vectors are commonly used to introduce genetic material into cells to modify cell function for a variety of purposes. Manufacture of those modified viruses may use a variety of cell types to generate high titers of viral particles; one of the most common being HEK293 cells. These cells have been modified into different lines aimed at satisfying specific use cases. HEK293T cells, for example, have been modified to include the SV40 large T antigen. Efficient viral particle production by HEK293T cells requires the maintenance of favorable cell culture conditions during expansion and transfection. This protocol describes the use of the Quantum® hollow-fiber bioreactor (HFB) system for the automated expansion of HEK293T cells, and the results derived using the protocol described herein were not compared with those from tissue culture flasks or other expansion platforms, as the parameters described are unique to Quantum's hollow fiber cell expansion environment. The purpose of this protocol is to help users of Quantum to focus on relevant parameters of expansion in the HFB milieu and to provide guidelines for a successful expansion of HEK293T cells in the Quantum system. The steps provided have been optimized to reliably control environmental factors related to glucose, lactate, and pH. Data reflecting this consistency are provided along with procedural time points reflected in text and figure formats.

Automated behavioral scoring: Do we even need humans?

The development of automated behavior scoring technology has been a tremendous boon to the study of social behavior. However, completely outsourcing behavioral analysis to a computer runs the risk of overlooking important nuances, and researchers risk distancing themselves from their very object of study. Here, I make the case that while automating analysis has been valuable, and overautomating analysis is risky, more effort should be spent automating the collection of behavioral data. Continuous automated behavioral observations conducted in situ have the promise to reduce confounding elements of social behavior research, such as handling stress, novel environments, one-time "snapshot" measures, and experimenter presence. Now that we have the capability to automatically process behavioral observations thanks to machine vision and machine learning, we would do well to leverage the same open-source ethos to increase the throughput of behavioral observation and collection. Fortunately, several such platforms have recently been developed. Repeated testing in the home environment will produce higher qualities and quantities of data, bringing us closer to realizing the ethological goals of studying animal behavior in a naturalistic context.

End-to-End Semi-automated Mid-scale Protein Screening Platform for Drug Discovery Research

The drug discovery landscape is ever-evolving and constantly demands revolutionary technology advancements in protein expression and production laboratories. We have built a higher-throughput mid-scale semi-automated protein expression and screening platform to accelerate drug discovery research. The workflow described here enables comprehensive expression and purification screening assessment of challenging or difficult-to-express recombinant proteins in a fast and efficient manner by delivering small but sufficient amounts of high-quality proteins. The platform has been implemented for a wide range of applications that include identification of optimal constructs and chaperones for poorly expressing proteins, assessment of co-expression partners for expressing stable multiprotein complexes, and suitable buffer/additive screening for insoluble or aggregation-prone proteins. The approach allows parallel expression, purification, and characterization of 24 different samples using co-infection or a polycistronic approach in insect cells and enables parallel testing of multiple parameters to improve protein yields. The strategy has been successfully adopted for screening intracellular and secreted proteins in Escherichia coli, mammalian transient expression, and baculovirus expression vector systems. Proteins purified from this platform are used for several structural and functional screens, such as negative staining, biochemical activity assays, mass spectrometry, surface plasmon resonance, and DNA-encoded chemical library screens. In this article, for simplicity, we have focused on detailed expression and purification screening of intracellular protein complexes from insect cells. © 2023 Wiley Periodicals LLC. Basic Protocol 1: Baculovirus generation via homologous recombination Support Protocol 1: Anti-glycoprotein 64 antibody assay Basic Protocol 2: Generation of insect cell biomass expressing target protein(s) Basic Protocol 3: Mid-scale affinity purification Support Protocol 2: Automated method for affinity purification on Hamilton STAR Basic Protocol 4: Size exclusion chromatography Support Protocol 3: Chromeleon 7 operation on Vanquish Duo.

An artificial intelligence-based pipeline for automated detection and localisation of epileptic sources from magnetoencephalography

Objective.Magnetoencephalography (MEG) is a powerful non-invasive diagnostic modality for presurgical epilepsy evaluation. However, the clinical utility of MEG mapping for localising epileptic foci is limited by its low efficiency, high labour requirements, and considerable interoperator variability. To address these obstacles, we proposed a novel artificial intelligence-based automated magnetic source imaging (AMSI) pipeline for automated detection and localisation of epileptic sources from MEG data.Approach.To expedite the analysis of clinical MEG data from patients with epilepsy and reduce human bias, we developed an autolabelling method, a deep-learning model based on convolutional neural networks and a hierarchical clustering method based on a perceptual hash algorithm, to enable the coregistration of MEG and magnetic resonance imaging, the detection and clustering of epileptic activity, and the localisation of epileptic sources in a highly automated manner. We tested the capability of the AMSI pipeline by assessing MEG data from 48 epilepsy patients.Main results.The AMSI pipeline was able to rapidly detect interictal epileptiform discharges with 93.31% ± 3.87% precision based on a 35-patient dataset (with sevenfold patientwise cross-validation) and robustly rendered accurate localisation of epileptic activity with a lobar concordance of 87.18% against interictal and ictal stereo-electroencephalography findings in a 13-patient dataset. We also showed that the AMSI pipeline accomplishes the necessary processes and delivers objective results within a much shorter time frame (∼12 min) than traditional manual processes (∼4 h).Significance.The AMSI pipeline promises to facilitate increased utilisation of MEG data in the clinical analysis of patients with epilepsy.

Perceived Use Cases, Barriers, and Requirements for a Smart Health-Tracking Toilet Seat: Qualitative Focus Group Study

BACKGROUND

Smart bathroom technology offers unrivaled opportunities for the automated measurement of a range of biomarkers and other data. Unfortunately, efforts in this area are mostly driven by a technology push rather than market pull approach, which decreases the chances of successful adoption. As yet, little is known about the use cases, barriers, and desires that potential users of smart bathrooms perceive.

OBJECTIVE

This study aimed to investigate how participants from the general population experience using a smart sensor-equipped toilet seat installed in their home. The study contributes to answering the following questions: What use cases do citizens see for this innovation? and What are the limitations and barriers to its everyday use that they see, including concerns regarding privacy, the lack of fit with everyday practices, and unmet expectations for user experience?

METHODS

Overall, 31 participants from 30 households participated in a study consisting of 3 (partially overlapping) stages: sensitizing, in which participants filled out questionnaires to trigger their thoughts about smart bathroom use and personal health; provotyping, in which participants received a gentle provocation in the form of a smart toilet seat, which they used for 2 weeks; and discussion, in which participants took part in a web-based focus group session to discuss their experiences.

RESULTS

Participants mostly found the everyday use of the toilet, including installation and dismantling when necessary, to be relatively easy and free of complications. Where complications occurred, participants mentioned issues related to the design of the prototype, technology, or mismatches with normal practices in using toilets and hygiene. A broad range of use cases were mentioned, ranging from signaling potentially detrimental health conditions or exacerbations of existing conditions to documenting physical data to measuring biomarkers to inform a diagnosis and behavioral change. Participants differed greatly in whether they let others use, or even know about, the seat. Ownership and control over their own data were essential for most participants.

CONCLUSIONS

This study showed that participants felt that a smart toilet seat could be acceptable and effective, as long as it fits everyday practices concerning toilet use and hygiene. The range of potential uses for a smart toilet seat is broad, as long as privacy and control over disclosure and data are warranted.

Automated analysis of written language in the three variants of primary progressive aphasia

Despite the important role of written language in everyday life, abnormalities in functional written communication have been sparsely investigated in primary progressive aphasia. Prior studies have analysed written language separately in each of the three variants of primary progressive aphasia-but have rarely compared them to each other or to spoken language. Manual analysis of written language can be a time-consuming process. We therefore developed a program that quantifies content units and total units in written or transcribed language samples. We analysed written and spoken descriptions of the Western Aphasia Battery picnic scene, based on a predefined content unit corpus. We calculated the ratio of content units to units as a measure of content density. Our cohort included 115 participants (20 controls for written, 20 controls for spoken, 28 participants with nonfluent variant primary progressive aphasia, 30 for logopenic variant and 17 for semantic variant). Our program identified content units with a validity of 99.7% (95%CI 99.5-99.8). All patients wrote fewer units than controls (P < 0.001). Patients with the logopenic variant (P = 0.013) and the semantic variant (0.004) wrote fewer content units than controls. The content unit-to-unit ratio was higher in the nonfluent and semantic variants than controls (P = 0.019), but no difference in the logopenic variant (P = 0.962). Participants with the logopenic (P < 0.001) and semantic (P = 0.04) variants produced fewer content units in written compared to spoken descriptions. All variants produced fewer units in written samples compared to spoken (P < 0.001). However, due to a relatively smaller decrease in written content units, we observed a larger content unit-to-unit ratio in writing over speech (P < 0.001). Written and spoken content units (r = 0.5, P = 0.009) and total units (r = 0.64, P < 0.001) were significantly correlated in patients with nonfluent variant, but this was not the case for logopenic or semantic. Considering all patients with primary progressive aphasia, fewer content units were produced in those with greater aphasia severity (Progressive Aphasia Severity Scale Sum of Boxes, r = -0.24, P = 0.04) and dementia severity (Clinical Dementia Rating scale Sum of Boxes, r = -0.34, P = 0.004). In conclusion, we observed reduced written content in patients with primary progressive aphasia compared to controls, with a preference for content over non-content units in patients with the nonfluent and semantic variants. We observed a similar 'telegraphic' style in both language modalities in patients with the nonfluent variant. Lastly, we show how our program provides a time-efficient tool, which could enable feedback and tracking of writing as an important feature of language and cognition.

Development of a Smart Portable Hypoxic Chamber with Accurate Sensing, Control and Visualization of In Vitro Cell Culture for Replication of Cancer Microenvironment

Clinical resistance towards treatment is a major concern in cancer therapy. This is due to in vitro studies lacking essential microenvironmental aspects. Tumor-hypoxia is an important pathophysiological phenomenon in numerous malignant tumors. Various studies have shown the importance of a hypoxic microenvironment (HME) in cancer drug resistance and its effects on cellular signaling and metabolism pathways. Most drugs fail in transition from a laboratory to clinical trials because of the variability in the testing microenvironment conditions. It is, thus, very crucial that research work needs to replicate these conditions in vitro to test the drugs and/or drug carriers for cancer therapy. Previous works have used a portable hypoxia chamber to reduce the cell microenvironment to hypoxic conditions. These techniques lack reliability and consistency due to a lack of control and visualization. In this research, we developed a smart portable hypoxia chamber that could accurately control the oxygen inside the portable chamber and have a global visualization. The proposed hypoxia chamber provided ease of use with the ranges of 1% to 20% oxygen with increments of 0.5%, as well as reproducibility and accuracy. The chamber displayed great precision on reaching the set oxygen limit and a high stability in maintaining that set level of oxygen compared to the uncontrolled setup for extended durations (24 h). For instance, at a 2% oxygen level, our automated system maintained this level over 1400 min, whereas the oxygen level fluctuated up to 4.5% in the conventional hypoxic chamber. We have also demonstrated the pitfalls of uncontrolled and non-visualized hypoxia chamber setup and the dire need for our system. The hypoxia-induced factor (HIF-1α) expression in cancer cell lines was tested and compared between the conventional hypoxia setup and our automated hypoxia chamber. We observed that there was a twofold increase in HIF-1α expression in the automated controlled chamber compared to the conventional device. The device also provided real-time sensing, visualization and control of the chamber conditions, which could aid in complex in vitro studies.

Vision transformer assisting rheumatologists in screening for capillaroscopy changes in systemic sclerosis: an artificial intelligence model

OBJECTIVES

The first objective of this study was to implement and assess the performance and reliability of a vision transformer (ViT)-based deep-learning model, an 'off-the-shelf' artificial intelligence solution, for identifying distinct signs of microangiopathy in nailfold capilloroscopy (NFC) images of patients with SSc. The second objective was to compare the ViT's analysis performance with that of practising rheumatologists.

METHODS

NFC images of patients prospectively enrolled in our European Scleroderma Trials and Research group (EUSTAR) and Very Early Diagnosis of Systemic Sclerosis (VEDOSS) local registries were used. The primary outcome investigated was the ViT's classification performance for identifying disease-associated changes (enlarged capillaries, giant capillaries, capillary loss, microhaemorrhages) and the presence of the scleroderma pattern in these images using a cross-fold validation setting. The secondary outcome involved a comparison of the ViT's performance vs that of rheumatologists on a reliability set, consisting of a subset of 464 NFC images with majority vote-derived ground-truth labels.

RESULTS

We analysed 17 126 NFC images derived from 234 EUSTAR and 55 VEDOSS patients. The ViT had good performance in identifying the various microangiopathic changes in capillaries by NFC [area under the curve (AUC) from 81.8% to 84.5%]. In the reliability set, the rheumatologists reached a higher average accuracy, as well as a better trade-off between sensitivity and specificity compared with the ViT. However, the annotators' performance was variable, and one out of four rheumatologists showed equal or lower classification measures compared with the ViT.

CONCLUSIONS

The ViT is a modern, well-performing and readily available tool for assessing patterns of microangiopathy on NFC images, and it may assist rheumatologists in generating consistent and high-quality NFC reports; however, the final diagnosis of a scleroderma pattern in any individual case needs the judgement of an experienced observer.

Automated Enrichment of Phosphotyrosine Peptides for High-Throughput Proteomics

Phosphotyrosine (pY) enrichment is critical for expanding the fundamental and clinical understanding of cellular signaling by mass spectrometry-based proteomics. However, current pY enrichment methods exhibit a high cost per sample and limited reproducibility due to expensive affinity reagents and manual processing. We present rapid-robotic phosphotyrosine proteomics (R2-pY), which uses a magnetic particle processor and pY superbinders or antibodies. R2-pY can handle up to 96 samples in parallel, requires 2 days to go from cell lysate to mass spectrometry injections, and results in global proteomic, phosphoproteomic, and tyrosine-specific phosphoproteomic samples. We benchmark the method on HeLa cells stimulated with pervanadate and serum and report over 4000 unique pY sites from 1 mg of peptide input, strong reproducibility between replicates, and phosphopeptide enrichment efficiencies above 99%. R2-pY extends our previously reported R2-P2 proteomic and global phosphoproteomic sample preparation framework, opening the door to large-scale studies of pY signaling in concert with global proteome and phosphoproteome profiling.

The experience of laser light feedback in back-squat resistance training

Introduction

The purpose of this paper is to contribute to the existing literature on performance in resistance training (RT) by addressing how a phenomenological perspective on experiences with inter kinaesthetic affectivity can illuminate experience of practicing RT with non-verbal, visual feedback provided through laser lights attached to the barbell.

Method

The material is created from qualitative interviews and using inter-kinaesthetic affectivity as analytical lenses.

Results

The findings show how participants interpret the feedback in the moment and explain how they adjust their movement in dialogue with the feedback and enable the "uptake" of feedback in their embodied experience. The findings show how the participants developed an awareness of how they can equalize the balance on their feet.

Discussion

We discuss what this means for the understanding of the training process in terms of how practitioners can use the uptake of non-verbal, visual feedback to immediately adjust the quality of their performance by responding kinaesthetically and bodily. The discussion contributes to the question of what kind of role a practitioner's own kinaesthetic and bodily experiences have in the development and organization of RT. Perspectives that include the lived and intersubjective body as a knowledge position are promising for illuminating the whole bodied engagement that is necessary to understand how to perform RT.

Automated analysis for multiplet identification from ultra-high resolution 2D- 1H, 13C-HSQC NMR spectra

Background: Metabolism is essential for cell survival and proliferation. A deep understanding of the metabolic network and its regulatory processes is often vital to understand and overcome disease. Stable isotope tracing of metabolism using nuclear magnetic resonance (NMR) and mass spectrometry (MS) is a powerful tool to derive mechanistic information of metabolic network activity. However, to retrieve meaningful information, automated tools are urgently needed to analyse these complex spectra and eliminate the bias introduced by manual analysis. Here, we present a data-driven algorithm to automatically annotate and analyse NMR signal multiplets in 2D- ¹H, ¹³C-HSQC NMR spectra arising from ¹³C - ¹³C scalar couplings. The algorithm minimises the need for user input to guide the analysis of 2D- ¹H, ¹³C-HSQC NMR spectra by performing automated peak picking and multiplet analysis. This enables non-NMR specialists to use this technology. The algorithm has been integrated into the existing MetaboLab software package. Methods: To evaluate the algorithm performance two criteria are tested: is the peak correctly annotated and secondly how confident is the algorithm with its analysis. For the latter a coefficient of determination is introduced. Three datasets were used for testing. The first was to test reproducibility with three biological replicates, the second tested the robustness of the algorithm for different amounts of scaling of the apparent J-coupling constants and the third focused on different sampling amounts. Results: The algorithm annotated overall >90% of NMR signals correctly with average coefficient of determination ρ of 94.06 ± 5.08%, 95.47 ± 7.20% and 80.47 ± 20.98% respectively. Conclusions: Our results indicate that the proposed algorithm accurately identifies and analyses NMR signal multiplets in ultra-high resolution 2D- ¹H, ¹³C-HSQC NMR spectra. It is robust to signal splitting enhancement and up to 25% of non-uniform sampling.

161Tb-DOTATOC Production Using a Fully Automated Disposable Cassette System: A First Step Toward the Introduction of 161Tb into the Clinic

¹⁶¹Tb is an interesting radionuclide for application in the treatment of neuroendocrine neoplasms' small metastases and single cancer cells because of its conversion and Auger-electron emission. Tb has coordination chemistry similar to that of Lu; therefore, like ¹⁷⁷Lu, it can stably radiolabel DOTATOC, one of the leading peptides used for the treatment of neuroendocrine neoplasms. However, ¹⁶¹Tb is a recently developed radionuclide that has not yet been specified for clinical use. Therefore, the aim of the current work was to characterize and specify ¹⁶¹Tb and to develop a protocol for the synthesis and quality control of ¹⁶¹Tb-DOTATOC with a fully automated process conforming to good-manufacturing-practice guidelines, in view of its clinical use. Methods: ¹⁶¹Tb, produced by neutron irradiation of ¹⁶⁰Gd in high-flux reactors followed by radiochemical separation from its target material, was characterized regarding its radionuclidic purity, chemical purity, endotoxin level, and radiochemical purity (RCP) in analogy to what is described in the European Pharmacopoeia for no-carrier-added ¹⁷⁷Lu. In addition, ¹⁶¹Tb was introduced into a fully automated cassette-module synthesis to produce ¹⁶¹Tb-DOTATOC, as used for ¹⁷⁷Lu-DOTATOC. The quality and stability of the produced radiopharmaceutical in terms of identity, RCP, and ethanol and endotoxin content were assessed by means of high-performance liquid chromatography, gas chromatography, and an endotoxin test, respectively. Results: ¹⁶¹Tb produced under the described conditions showed, as the no-carrier-added ¹⁷⁷Lu, a pH of 1-2, radionuclidic purity and RCP of more than 99.9%, and an endotoxin level below the permitted range (175 IU/mL), indicating its appropriate quality for clinical use. In addition, an efficient and robust procedure for the automated production and quality control of ¹⁶¹Tb-DOTATOC with clinically applicable specifications and activity levels, that is, 1.0-7.4 GBq in 20 mL, was developed. The radiopharmaceutical's quality control was also developed using chromatographic methods, which confirmed the product's stability (RCP ≥ 95%) over 24 h. Conclusion: The current study demonstrated that ¹⁶¹Tb has appropriate features for clinical use. The developed synthesis protocol guarantees high yields and safe preparation of injectable ¹⁶¹Tb-DOTATOC. The investigated approach could be translated to other DOTA-derivatized peptides; thus, ¹⁶¹Tb could be successfully applied in clinical practice for radionuclide therapy.

Just preparation for war and AI-enabled weapons

This paper maintains that the just war tradition provides a useful framework for analyzing ethical issues related to the development of weapons that incorporate artificial intelligence (AI), or "AI-enabled weapons." While development of any weapon carries the risk of violations of jus ad bellum and jus in bello, AI-enabled weapons can pose distinctive risks of these violations. The article argues that developing AI-enabled weapons in accordance with jus ante bellum principles of just preparation for war can help minimize the risk of these violations. These principles impose two obligations. The first is that before deploying an AI-enabled weapon a state must rigorously test its safety and reliability, and conduct review of its ability to comply with international law. Second, a state must develop AI-enabled weapons in ways that minimize the likelihood that a security dilemma will arise, in which other states feel threatened by this development and hasten to deploy such weapons without sufficient testing and review. Ethical development of weapons that incorporate AI therefore requires that a state focus not only on its own activity, but on how that activity is perceived by other states.

Characterization of an automated method to segment the human locus coeruleus

Following the development of magnetic resonance imaging (MRI) methods to assay the integrity of catecholamine nuclei, including the locus coeruleus (LC), there has been an effort to develop automated methods that can accurately segment this small structure in an automated manner to promote its widespread use and overcome limitations of manual segmentation. Here we characterize an automated LC segmentation approach (referred to as the funnel-tip [FT] method) in healthy individuals and individuals with LC degeneration in the context of Alzheimer's disease (AD, confirmed with tau-PET imaging using [18F]MK6240). The first sample included n = 190 individuals across the AD spectrum from cognitively normal to moderate AD. LC signal assayed with FT segmentation showed excellent agreement with manual segmentation (intraclass correlation coefficient [ICC] = 0.91). Compared to other methods, the FT method showed numerically higher correlation to AD status (defined by presence of tau: Cohen's d = 0.64) and AD severity (Braak stage: Pearson R = -.35, cognitive function: R = .25). In a separate sample of n = 12 control participants, the FT method showed excellent scan-rescan reliability (ICC = 0.82). In another sample of n = 30 control participants, we found that the structure of the LC defined by FT segmentation approximated its expected shape as a contiguous line: <5% of LC voxels strayed >1 voxel (0.69 mm) from this line. The FT LC segmentation shows high agreement with manual segmentation and captures LC degeneration in AD. This practical method may facilitate larger research studies of the human LC-norepinephrine system and has potential to support future use of neuromelanin-sensitive MRI as a clinical biomarker.

Development of a Novel, Automated High-Throughput Device for Performing the Comet Assay

A comet assay is a trusted and widely used method for assessing DNA damage in individual eukaryotic cells. However, it is time-consuming and requires extensive monitoring and sample manipulation by the user. This limits the throughput of the assay, increases the risk of errors, and contributes to intra- and inter-laboratory variability. Here, we describe the development of a device which automates high throughput sample processing for a comet assay. This device is based upon our patented, high throughput, vertical comet assay electrophoresis tank, and incorporates our novel, patented combination of assay fluidics, temperature control, and a sliding electrophoresis tank to facilitate sample loading and removal. Additionally, we demonstrated that the automated device performs at least as well as our "manual" high throughput system, but with all the advantages of a fully "walkaway" device, such as a decreased need for human involvement and a decreased assay run time. Our automated device represents a valuable, high throughput approach for reliably assessing DNA damage with the minimal operator involvement, particularly if combined with the automated analysis of comets.

Comparative Evaluation between Visual and Automated Dipstick Urinalyses in Dogs

Urine test strips are commercially available and can be assessed with semi-automated analyzers or by visual assessment. This study aimed to compare the visual and automated evaluations of dipstick variables in canine urine samples. One hundred and nineteen urine samples were evaluated. Automated analysis was performed on a veterinary urine analyzer URIT-50Vet (URIT Medical Electronic) with UC VET13 Plus strips. Multistix 10 SG dipsticks (Siemens Healthcare GmbH, Erlangen, Germany) were used for visual evaluation, along with a refractometer (Clinical Refractometer Atago T2-Ne, Atago Co., Tokyo, Japan) for urine specific gravity measurements. A linear relationship was observed between the pH measurements (p = 0.2) of the two methods; the Passing-Bablok procedure was valid since neither proportional nor systematic significant errors were observed. Comparing the two methods, the correlation for urine specific gravity was poor (p = 0.01, CI 0.667-1.000). Moderate agreement was demonstrated for proteins (κ = 0.431), bilirubin (κ = 0.434) and glucose (κ = 0.450). Agreement was substantial for blood (κ = 0.620) and poor for leukocytes (κ = 0.100). Poor agreement was observed for ketones (κ = -0.006). Apart from the pH analysis, visual and automated dipstick urinalyses should not be used interchangeably. Multiple urine samples obtained from the same dog during the day should be evaluated using the same method to overcome erroneous results.

Towards Automated and High-Throughput Quantitative Sizing and Isotopic Analysis of Nanoparticles via Single Particle-ICP-TOF-MS

The work described herein assesses the ability to characterize gold nanoparticles (Au NPs) of 50 and 100 nm, as well as 60 nm silver shelled gold core nanospheres (Au/Ag NPs), for their mass, respective size, and isotopic composition in an automated and unattended fashion. Here, an innovative autosampler was employed to mix and transport the blanks, standards, and samples into a high-efficiency single particle (SP) introduction system for subsequent analysis by inductively coupled plasma-time of flight-mass spectrometry (ICP-TOF-MS). Optimized NP transport efficiency into the ICP-TOF-MS was determined to be >80%. This combination, SP-ICP-TOF-MS, allowed for high-throughput sample analysis. Specifically, 50 total samples (including blanks/standards) were analyzed over 8 h, to provide an accurate characterization of the NPs. This methodology was implemented over the course of 5 days to assess its long-term reproducibility. Impressively, the in-run and day-to-day variation of sample transport is assessed to be 3.54 and 9.52% relative standard deviation (%RSD), respectively. The determination of Au NP size and concentration was of <5% relative difference from the certified values over these time periods. Isotopic characterization of the ¹⁰⁷Ag/¹⁰⁹Ag particles (n = 132,630) over the course of the measurements was determined to be 1.0788 ± 0.0030 with high accuracy (0.23% relative difference) when compared to the multi-collector-ICP-MS determination.

Longitudinal home-cage automated assessment of climbing behavior shows sexual dimorphism and aging-related decrease in C57BL/6J healthy mice and allows early detection of motor impairment in the N171-82Q mouse model of Huntington's disease

Monitoring the activity of mice within their home cage is proving to be a powerful tool for revealing subtle and early-onset phenotypes in mouse models. Video-tracking, in particular, lends itself to automated machine-learning technologies that have the potential to improve the manual annotations carried out by humans. This type of recording and analysis is particularly powerful in objective phenotyping, monitoring behaviors with no experimenter intervention. Automated home-cage testing allows the recording of non-evoked voluntary behaviors, which do not require any contact with the animal or exposure to specialist equipment. By avoiding stress deriving from handling, this approach, on the one hand, increases the welfare of experimental animals and, on the other hand, increases the reliability of results excluding confounding effects of stress on behavior. In this study, we show that the monitoring of climbing on the wire cage lid of a standard individually ventilated cage (IVC) yields reproducible data reflecting complex phenotypes of individual mouse inbred strains and of a widely used model of neurodegeneration, the N171-82Q mouse model of Huntington's disease (HD). Measurements in the home-cage environment allowed for the collection of comprehensive motor activity data, which revealed sexual dimorphism, daily biphasic changes, and aging-related decrease in healthy C57BL/6J mice. Furthermore, home-cage recording of climbing allowed early detection of motor impairment in the N171-82Q HD mouse model. Integrating cage-floor activity with cage-lid activity (climbing) has the potential to greatly enhance the characterization of mouse strains, detecting early and subtle signs of disease and increasing reproducibility in preclinical studies.

Changes in Radiology Due to Artificial Intelligence That Can Attract Medical Students to the Specialty

The role of artificial intelligence (AI) in radiology has grown exponentially in the recent years. One of the primary worries by medical students is that AI will cause the roles of a radiologist to become automated and thus obsolete. Therefore, there is a greater hesitancy by medical students to choose radiology as a specialty. However, it is in this time of change that the specialty needs new thinkers and leaders. In this succinct viewpoint, 2 medical students involved in AI and 2 radiologists specializing in AI or clinical informatics posit that not only are these fears false, but the field of radiology will be transformed in such a way due to AI that there will be novel reasons to choose radiology. These new factors include greater impact on patient care, new space for innovation, interdisciplinary collaboration, increased patient contact, becoming master diagnosticians, and greater opportunity for global health initiatives, among others. Finally, since medical students view mentorship as a critical resource when deciding their career path, medical educators must also be cognizant of these changes and not give much credence to the prevalent fearmongering. As the field and practice of radiology continue to undergo significant change due to AI, it is urgent and necessary for the conversation to expand from expert to expert to expert to student. Medical students should be encouraged to choose radiology specifically because of the changes brought on by AI rather than being deterred by it.

Automating Case Reporting of Chlamydia and Gonorrhea to Public Health Authorities in Illinois Clinics: Implementation and Evaluation of Findings

BACKGROUND

Chlamydia and gonorrhea cases continue to rise in Illinois, increasing by 16.4% and 70.9% in 2019, respectively, compared with 2015. Providers are required to report both chlamydia and gonorrhea, as mandated by public health laws. Manual reporting remains a huge burden; 90%-93% of cases were reported to Illinois Department of Public Health (IDPH) via electronic laboratory reporting (ELR), and the remaining were reported through web-based data entry platforms, faxes, and phone calls. However, cases reported via ELRs only contain information available to a laboratory facility and do not contain additional data needed for public health. Such data are typically found in an electronic health record (EHR). Electronic case reports (eCRs) were developed and automated the generation of case reports from EHRs to be reported to public health agencies.

OBJECTIVE

Prior studies consolidated trigger criteria for eCRs, and compared with manual reporting, found it to be more complete. The goal of this project is to pilot standards-based eCR for chlamydia and gonorrhea. We evaluated the throughput, completeness, and timeliness of eCR compared to ELR, as well as the implementation experience at a large health center-controlled network in Illinois.

METHODS

For this study, we selected 8 clinics located on the north, west, and south sides of Chicago to implement the eCRs; these cases were reported to IDPH. The study period was 52 days. The centralized EHR used by these clinics leveraged 2 of the 3 case detection scenarios, which were previously defined as the trigger, to generate an eCR. These messages were successfully transmitted via Health Level 7 electronic initial case report standard. Upon receipt by IDPH, these eCRs were parsed and housed in a staging database.

RESULTS

During the study period, 183 eCRs representing 135 unique patients were received by IDPH. eCR reported 95% (n=113 cases) of all the chlamydia cases and 97% (n=70 cases) of all the gonorrhea cases reported from the participating clinical sites. eCR found an additional 14 (19%) cases of gonorrhea that were not reported via ELR. However, ELR reported an additional 6 cases of chlamydia and 2 cases of gonorrhea, which were not reported via eCR. ELR reported 100% of chlamydia cases but only 81% of gonorrhea cases. While key elements such as patient and provider names were complete in both eCR and ELR, eCR was found to report additional clinical data, including history of present illness, reason for visit, symptoms, diagnosis, and medications.

CONCLUSIONS

eCR successfully identified and created automated reports for chlamydia and gonorrhea cases in the implementing clinics in Illinois. eCR demonstrated a more complete case report and represents a promising future of reducing provider burden for reporting cases while achieving greater semantic interoperability between health care systems and public health.

Proposal of a Method for Transferring High-Quality Scientific Literature Data to Virtual Patient Cases Using Categorical Data Generated by Bernoulli-Distributed Random Values: Development and Prototypical Implementation

BACKGROUND

Teaching medicine is a complex task because medical teachers are also involved in clinical practice and research and the availability of cases with rare diseases is very restricted. Automatic creation of virtual patient cases would be a great benefit, saving time and providing a wider choice of virtual patient cases for student training.

OBJECTIVE

This study explored whether the medical literature provides usable quantifiable information on rare diseases. The study implemented a computerized method that simulates basic clinical patient cases utilizing probabilities of symptom occurrence for a disease.

METHODS

Medical literature was searched for suitable rare diseases and the required information on the respective probabilities of specific symptoms. We developed a statistical script that delivers basic virtual patient cases with random symptom complexes generated by Bernoulli experiments, according to probabilities reported in the literature. The number of runs and thus the number of patient cases generated are arbitrary.

RESULTS

We illustrated the function of our generator with the exemplary diagnosis "brain abscess" with the related symptoms "headache, mental status change, focal neurologic deficit, fever, seizure, nausea and vomiting, nuchal rigidity, and papilledema" and the respective probabilities from the literature. With a growing number of repetitions of the Bernoulli experiment, the relative frequencies of occurrence increasingly converged with the probabilities from the literature. For example, the relative frequency for headache after 10.000 repetitions was 0.7267 and, after rounding, equaled the mean value of the probability range of 0.73 reported in the literature. The same applied to the other symptoms.

CONCLUSIONS

The medical literature provides specific information on characteristics of rare diseases that can be transferred to probabilities. The results of our computerized method suggest that automated creation of virtual patient cases based on these probabilities is possible. With additional information provided in the literature, an extension of the generator can be implemented in further research.

Automated Insulin Delivery Systems as a Treatment for Type 2 Diabetes Mellitus: A Review

OBJECTIVE

Approximately 6.3% of the worldwide population has type 2 diabetes mellitus (T2DM), and the number of people requiring insulin is increasing. Automated insulin delivery (AID) systems integrate continuous subcutaneous insulin infusion and continuous glucose monitoring with a predictive control algorithm to provide more physiologic glycemic control. Personalized glycemic targets are recommended in T2DM owing to the heterogeneity of the disease. Based on the success of hybrid closed-loop systems in improving glycemic control and safety in type 1 diabetes mellitus, there has been further interest in the use of these systems in people with T2DM.

METHODS

We performed a review of AID systems with a focus on the T2DM population.

RESULTS

In 5 randomized controlled trials, AID systems improve time in range and reduce glycemic variability, without increasing insulin requirements or the risk of hypoglycemia.

CONCLUSION

AID systems in T2DM are safe and effective in hospitalized and closely monitored settings. Home studies of longer duration are required to assess for long-term benefit and identify target populations of benefit.

Automated segmentation of the healed anterior cruciate ligament from T2 * relaxometry MRI scans

Collagen organization of the anterior cruciate ligament (ACL) can be evaluated using T2 * relaxometry. However, T2 * mapping requires manual image segmentation, which is a time-consuming process and prone to inter- and intra- segmenter variability. Automating segmentation would address these challenges. A model previously trained using Constructive Interference in Steady State (CISS) scans was applied to T2 * segmentation via transfer learning. It was hypothesized that there would be no significant differences in the model's segmentation performance between T2 * and CISS, structural measures versus ground truth manual segmentation, and reliability versus independent and retest manual segmentation. Transfer learning was conducted using 54 T2 * scans of the ACL. Segmentation performance was assessed with Dice coefficient, precision, and sensitivity, and structurally with T2 * value, volume, subvolume proportions, and cross-sectional area. Model performance relative to independent manual segmentation and repeated segmentation by the ground truth segmenter (retest) were evaluated on a random subset. Segmentation performance was analyzed with Mann-Whitney U tests, structural measures with Wilcoxon signed-rank tests, and performance relative to manual segmentation with repeated-measures analysis of variance/Tukey tests (α = 0.05). T2 * segmentation performance was not significantly different from CISS on all measures (p > 0.35). No significant differences were detected in structural measures (p > 0.50). Automatic segmentation performed as well as the retest on all segmentation measures, whereas independent segmentations were lower than retest and/or automatic segmentation (p < 0.023). Structural measures were not significantly different between segmenters. The automatic segmentation model performed as well on the T2 * sequence as on CISS and outperformed independent manual segmentation while performing as well as retest segmentation.

A Novel Tactile Function Assessment Using a Miniature Tactile Stimulator

Several methods for the measurement of tactile acuity have been devised previously, but unexpected nonspatial cues and intensive manual skill requirements compromise measurement accuracy. Therefore, we must urgently develop an automated, accurate, and noninvasive method for assessing tactile acuity. The present study develops a novel method applying a robotic tactile stimulator to automatically measure tactile acuity that comprises eye-opened, eye-closed training, and testing sessions. Healthy participants judge the orientation of a rotating grating ball presented on their index fingerpads in a two-alternative forced-choice task. A variable rotation speed of 5, 10, 40, or 160 mm/s was used for the tactile measurement at a variety of difficulties. All participants met the passing criteria for the training experiment. Performance in orientation identification, quantified by the proportion of trials with correct answers, differed across scanning directions, with the highest rotation speed (160 mm/s) having the worst performance. Accuracy did not differ between vertical and horizontal orientations. Our results demonstrated the utility of the pre-test training protocol and the functionality of the developed procedure for tactile acuity assessment. The novel protocol performed well when applied to the participants. Future studies will be conducted to apply this method to patients with impairment of light touch.

Evaluation of automated radiostereometric image registration in total knee arthroplasty utilizing a synthetic-based and a CT-based volumetric model

Radiostereometic analysis (RSA) is an accurate method for rigid body pose (position and orientation) in three-dimensional space. Traditionally, RSA is based on insertion of periprosthetic tantalum markers and manual implant contour selection which limit clinically application. We propose an automated image registration technique utilizing digitally reconstructed radiographs (DRR) of computed tomography (CT) volumetric bone models (autorsa-bone) as a substitute for tantalum markers. Furthermore, an automated synthetic volumetric representation of total knee arthroplasty implant models (autorsa-volume) to improve previous silhouette-projection methods (autorsa-surface). As reference, we investigated the accuracy of implanted tantalum markers (marker) or a conventional manually contour-based method (mbrsa) for the femur and tibia. The data are presented as mean (standard deviation). The autorsa-bone method displayed similar accuracy of -0.013 (0.075) mm compared to the gold standard method (marker) of -0.013 (0.085). The autorsa-volume with 0.034 (0.106) mm did not markedly improve the autorsa-surface with 0.002 (0.129) mm, and none of these reached the mbrsa method of -0.009 (0.094) mm. In conclusion, marker-free RSA is feasible with similar accuracy as gold standard utilizing DRR and CT obtained volumetric bone models. Furthermore, utilizing synthetic generated volumetric implant models could not improve the silhouette-based method. However, with a slight loss of accuracy the autorsa methods provide a feasible automated alternative to the semi-automated method.

Preclinical evaluation of semi-automated laser ablation for pulmonary vein isolation: A comparative study

INTRODUCTION

Visually-guided laser balloon ablation (VGLA) currently requires careful manual rotation of the laser to create overlapping lesions. A novel semi-automated VGLA may reduce ablation times and lesion gaps. We aimed to compare semi-automated (SA) VGLA to that of manual (MN) VGLA.

METHODS

Acute: Nine swine underwent right superior pulmonary vein isolation (PVI) using either SA (n = 3, 13-18 W), MN (n = 3, 8.5-12 W), or radiofrequency (RF, n = 3, 25-40 W) and were killed acutely. Chronic: 16 swine, underwent PVI using either SA (n = 8, 15 W) or MN (n = 8, 10 W), and were survived for 1 month before being killed. All hearts were then submitted for pathological evaluation.

RESULTS

Acute: PVI was successful in all 9/9 swine with lesion counts significantly lower in the SA arm (5.3 ± 5.9, 33.7 ± 10.0, and 28.0 ± 4.4 in SA, MN, and RF arms; p = .007 for SA and MN). At necropsy, circumferentiality and transmurality were 98% and 94% in SA, 98% and 80% in MN, and 100% and 100% in RF arms. A single steam pop was noted on sectioning in the SA arm swine and occurred in the high dose (18 W) strategy. Chronic: PVI was acutely successful in 16/16 swine with no difference in PVI durability rates (62.5% vs. 75.0%), lesion transmurality (95.8 ± 17.4% vs.91.9 ± 25.9%), and circumferentiality (95.8 ± 6.6% vs. 94.8 ± 6.3%) between SA and MN arms. Catheter use time and lesion counts were lower in the SA arm compared to the MN arm (11.5 ± 12.7 vs. 21.8 ± 3.8 min, p = .046 and 4.8 ± 3.83 vs. 35.4 ± 4.4, p < .001).

CONCLUSION

Motor-assisted semi-automated laser balloon ablation can improve upon procedural efficiency by reducing ablation time.

An Automated Microfluidic Analyzer for In Situ Monitoring of Total Alkalinity

We have designed, built, tested, and deployed an autonomous in situ analyzer for seawater total alkalinity. Such analyzers are required to understand the ocean carbon cycle, including anthropogenic carbon dioxide (CO2) uptake and for mitigation efforts via monitoring, reporting, and verification of carbon dioxide removal through ocean alkalinity enhancement. The microfluidic nature of our instrument makes it relatively lightweight, reagent efficient, and amenable for use on platforms that would carry it on long-term deployments. Our analyzer performs a series of onboard closed-cell titrations with three independent stepper-motor driven syringe pumps, providing highly accurate mixing ratios that can be systematically swept through a range of pH values. Temperature effects are characterized over the range 5-25 °C allowing for field use in most ocean environments. Each titration point requires approximately 170 μL of titrant, 830 μL of sample, 460 J of energy, and a total of 105 s for pumping and optical measurement. The analyzer performance is demonstrated through field data acquired at two sites, representing a cumulative 25 days of operation, and is evaluated against laboratory measurements of discrete water samples. Once calibrated against onboard certified reference material, the analyzer showed an accuracy of -0.17 ± 24 μmol kg-1. We further report a precision of 16 μmol kg-1, evaluated on repeated in situ measurements of the aforementioned certified reference material. The total alkalinity analyzer presented here will allow measurements to take place in remote areas over extended periods of time, facilitating affordable observations of a key parameter of the ocean carbon system with high spatial and temporal resolution.

High-throughput combined voltage-clamp/current-clamp analysis of freshly isolated neurons

The patch-clamp technique is the gold-standard methodology for analysis of excitable cells. However, throughput of manual patch-clamp is slow, and high-throughput robotic patch-clamp, while helpful for applications like drug screening, has been primarily used to study channels and receptors expressed in heterologous systems. We introduce an approach for automated high-throughput patch-clamping that enhances analysis of excitable cells at the channel and cellular levels. This involves dissociating and isolating neurons from intact tissues and patch-clamping using a robotic instrument, followed by using an open-source Python script for analysis and filtration. As a proof of concept, we apply this approach to investigate the biophysical properties of voltage-gated sodium (Nav) channels in dorsal root ganglion (DRG) neurons, which are among the most diverse and complex neuronal cells. Our approach enables voltage- and current-clamp recordings in the same cell, allowing unbiased, fast, simultaneous, and head-to-head electrophysiological recordings from a wide range of freshly isolated neurons without requiring culturing on coverslips.

A Safe Place to Learn: Peer Research Qualitative Investigation of gameChange Virtual Reality Therapy

BACKGROUND

Automated virtual reality (VR) therapy has the potential to substantially increase access to evidence-based psychological treatments. The results of a multicenter randomized controlled trial showed that gameChange VR cognitive therapy reduces the agoraphobic avoidance of people diagnosed with psychosis, especially for those with severe avoidance.

OBJECTIVE

We set out to use a peer research approach to explore participants' experiences with gameChange VR therapy. This in-depth experiential exploration of user experience may inform the implementation in clinical services and future VR therapy development.

METHODS

Peer-led semistructured remote interviews were conducted with 20 people with a diagnosis of psychosis who had received gameChange as part of the clinical trial (ISRCTN17308399). Data were analyzed using interpretative phenomenological analysis and template analyses. A multiperspectival approach was taken to explore subgroups. Credibility checks were conducted with the study Lived Experience Advisory Panel.

RESULTS

Participants reported the substantial impact of anxious avoidance on their lives before the VR intervention, leaving some of them housebound and isolated. Those who were struggling the most with agoraphobic avoidance expressed the most appreciation for, and gains from, the gameChange therapy. The VR scenarios provided "a place to practise." Immersion within the VR scenarios triggered anxiety, yet participants were able to observe this and respond in different ways than usual. The "security of knowing the VR scenarios are not real" created a safe place to learn about fears. The "balance of safety and anxiety" could be calibrated to the individual. The new learning made in VR was "taken into the real world" through practice and distilling key messages with support from the delivery staff member.

CONCLUSIONS

Automated VR can provide a therapeutic simulation that allows people diagnosed with psychosis to learn and embed new ways of responding to the situations that challenge them. An important process in anxiety reduction is enabling the presentation of stimuli that induce the original anxious fears yet allow for learning of safety. In gameChange, the interaction of anxiety and safety could be calibrated to provide a safe place to learn about fears and build confidence. This navigation of therapeutic learning can be successfully managed by patients themselves in an automated therapy, with staff support, that provides users with personalized control. The clinical improvements for people with severe anxious avoidance, the positive experience of VR, and the maintenance of a sense of control are likely to facilitate implementation.

Automated Enrichment of Phosphotyrosine Peptides for High-Throughput Proteomics

Phosphotyrosine (pY) enrichment is critical for expanding fundamental and clinical understanding of cellular signaling by mass spectrometry-based proteomics. However, current pY enrichment methods exhibit a high cost per sample and limited reproducibility due to expensive affinity reagents and manual processing. We present rapid-robotic phosphotyrosine proteomics (R2-pY), which uses a magnetic particle processor and pY superbinders or antibodies. R2-pY handles 96 samples in parallel, requires 2 days to go from cell lysate to mass spectrometry injections, and results in global proteomic, phosphoproteomic and tyrosine specific phosphoproteomic samples. We benchmark the method on HeLa cells stimulated with pervanadate and serum and report over 4000 unique pY sites from 1 mg of peptide input, strong reproducibility between replicates, and phosphopeptide enrichment efficiencies above 99%. R2-pY extends our previously reported R2-P2 proteomic and global phosphoproteomic sample preparation framework, opening the door to large-scale studies of pY signaling in concert with global proteome and phosphoproteome profiling.

Factors associated with the time and magnitude of the nadir body condition score in early lactation and its subsequent effects on fertility of Holstein cows

The objective of this study was to identify variables associated with the time and magnitude of nadir BCS (nBCS) during early lactation and to characterize the effects of these two measurements on the subsequent pregnancy at first artificial insemination (P/AI1) and pregnancy loss of Holstein cows. A retrospective observational study was completed using data collected from 12,042 lactations in 7,626 Holstein cows calving between April 2019 and January 2022 in a commercial dairy operation located in Colorado, USA. Scores generated by an automated BCS camera system at dry-off (BCSdry), calving (BCS1), nadir (nBCS; defined as the lowest daily BCS from calving to 100 DIM), and 21 d post nadir were selected for the analyses and subsequently categorized into quartiles (Q1 = lower nBCS). Changes in BCS were calculated from dry-off to calving (multiparous cows); from calving to nadir; and from nadir to 21 d post nadir and assigned into quartile categories considering Q1 as the 25% of cows with greater loss (pre nadir) or smaller gain (post nadir) of BCS. Data were examined using logistic regression, ANOVA, and time to event analyses. Initial univariable models were followed by multivariable models that considered parity category, calving season, BCS1, and average milk yield up to 30 DIM (M30) as covariables. Sick cows and high producing cows had increased DIM at nadir, while lower BCSdry, greater M30, and occurrence of disease were associated with lower nBCS. Cows that conceived at AI1 had their nBCS earlier (45.1 ± 0.32 d) than cows that remained open after first AI (46.4 ± 0.35 d). On the contrary, DIM at nBCS did not differ by pregnancy loss status. The logistic regression analyses reported smaller odds of P/AI1 in cows in the lower nBCS categories. Likewise, cows with larger loss in BCS between calving and nadir had smaller odds of P/AI1. Although cows that lost their pregnancy had lower nBCS, this association was not consistent across the analyses. The time to and magnitude of nBCS, together with the ∆BCS from calving to nadir could be useful measures to evaluate the transition and early lactation status of dairy cows.

A Smartphone-Based Platform Assisted by Artificial Intelligence for Reading and Reporting Rapid Diagnostic Tests: Evaluation Study in SARS-CoV-2 Lateral Flow Immunoassays

BACKGROUND

Rapid diagnostic tests (RDTs) are being widely used to manage COVID-19 pandemic. However, many results remain unreported or unconfirmed, altering a correct epidemiological surveillance.

OBJECTIVE

Our aim was to evaluate an artificial intelligence-based smartphone app, connected to a cloud web platform, to automatically and objectively read RDT results and assess its impact on COVID-19 pandemic management.

METHODS

Overall, 252 human sera were used to inoculate a total of 1165 RDTs for training and validation purposes. We then conducted two field studies to assess the performance on real-world scenarios by testing 172 antibody RDTs at two nursing homes and 96 antigen RDTs at one hospital emergency department.

RESULTS

Field studies demonstrated high levels of sensitivity (100%) and specificity (94.4%, CI 92.8%-96.1%) for reading IgG band of COVID-19 antibody RDTs compared to visual readings from health workers. Sensitivity of detecting IgM test bands was 100%, and specificity was 95.8% (CI 94.3%-97.3%). All COVID-19 antigen RDTs were correctly read by the app.

CONCLUSIONS

The proposed reading system is automatic, reducing variability and uncertainty associated with RDTs interpretation and can be used to read different RDT brands. The web platform serves as a real-time epidemiological tracking tool and facilitates reporting of positive RDTs to relevant health authorities.

Automated Kidney and Liver Segmentation in MR Images in Patients with Autosomal Dominant Polycystic Kidney Disease: A Multicenter Study

Background

Imaging-based total kidney volume (TKV) and total liver volume (TLV) are major prognostic factors in autosomal dominant polycystic kidney disease (ADPKD) and end points for clinical trials. However, volumetry is time consuming and reader dependent in clinical practice. Our aim was to develop a fully automated method for joint kidney and liver segmentation in magnetic resonance imaging (MRI) and to evaluate its performance in a multisequence, multicenter setting.

Methods

The convolutional neural network was trained on a large multicenter dataset consisting of 992 MRI scans of 327 patients. Manual segmentation delivered ground-truth labels. The model's performance was evaluated in a separate test dataset of 93 patients (350 MRI scans) as well as a heterogeneous external dataset of 831 MRI scans from 323 patients.

Results

The segmentation model yielded excellent performance, achieving a median per study Dice coefficient of 0.92-0.97 for the kidneys and 0.96 for the liver. Automatically computed TKV correlated highly with manual measurements (intraclass correlation coefficient [ICC]: 0.996-0.999) with low bias and high precision (-0.2%±4% for axial images and 0.5%±4% for coronal images). TLV estimation showed an ICC of 0.999 and bias/precision of -0.5%±3%. For the external dataset, the automated TKV demonstrated bias and precision of -1%±7%.

Conclusions

Our deep learning model enabled accurate segmentation of kidneys and liver and objective assessment of TKV and TLV. Importantly, this approach was validated with axial and coronal MRI scans from 40 different scanners, making implementation in clinical routine care feasible.Clinical Trial registry name and registration number: The German ADPKD Tolvaptan Treatment Registry (AD[H]PKD), NCT02497521.

degenerated articular cartilage in mice

Characterizing the biomechanical properties of articular cartilage is crucial to understanding processes of tissue homeostasis vs. degeneration. In mouse models, however, limitations are imposed by their small joint size and thin cartilage surfaces. Here we present a three-dimensional (3D) automated surface mapping system and methodology that allows for mechanical characterization of mouse cartilage with high spatial resolution. We performed repeated indentation mappings, followed by cartilage thickness measurement via needle probing, at 31 predefined positions distributed over the medial and lateral femoral condyles of healthy mice. High-resolution 3D x-ray microscopy (XRM) imaging was used to validate tissue thickness measurements. The automated indentation mapping was reproducible, and needle probing yielded cartilage thicknesses comparable to XRM imaging. When comparing healthy vs. degenerated cartilage, topographical variations in biomechanics were identified, with altered thickness and stiffness (instantaneous modulus) across condyles and within anteroposterior sub-regions. This quantitative technique comprehensively characterized cartilage function in mice femoral condyle cartilage. Hence, it has the potential to improve our understanding of tissue structure-function interplay in mouse models of repair and disease.

SpecSeg is a versatile toolbox that segments neurons and neurites in chronic calcium imaging datasets based on low-frequency cross-spectral power

Imaging calcium signals in neurons of animals using single- or multi-photon microscopy facilitates the study of coding in large neural populations. Such experiments produce massive datasets requiring powerful methods to extract responses from hundreds of neurons. We present SpecSeg, an open-source toolbox for (1) segmentation of regions of interest (ROIs) representing neuronal structures, (2) inspection and manual editing of ROIs, (3) neuropil correction and signal extraction, and (4) matching of ROIs in sequential recordings. ROI segmentation in SpecSeg is based on temporal cross-correlations of low-frequency components derived by Fourier analysis of each pixel with its neighbors. The approach is user-friendly, intuitive, and insightful and enables ROI detection around neurons or neurites. It works for single- (miniscope) and multi-photon microscopy data, eliminating the need for separate toolboxes. SpecSeg thus provides an efficient and versatile approach for analyzing calcium responses in neuronal structures imaged over prolonged periods of time.

Evaluation of MetaSystems Automated Fluorescent Microscopy System for the Machine-Assisted Detection of Acid-Fast Bacilli in Clinical Samples

Manual reading of fluorescent acid-fast bacilli (AFB) microscopy slides is time-intensive and technically demanding. The aim of this study was to evaluate the accuracy of MetaSystems' automated fluorescent AFB slide scanner and analyzer. Auramine O-stained slides corresponding to 133 culture-positive and 363 culture-negative respiratory (n = 284), tissue (n = 120), body fluid (n = 81), and other (n = 11) sources were evaluated with the MetaSystems Mycobacteria Scanner running the NEON Metafer AFB Module. The sensitivity and specificity of the MetaSystems platform was measured as a standalone diagnostic and as an assistant to technologists to review positive images. Culture results were used as the reference method. The MetaSystems platform failed to scan 57 (11.5%) slides. The MetaSystems platform used as a standalone had a sensitivity of 97.0% (129/133; 95% CI 92.5 to 99.2) and specificity of 12.7% (46/363; 95% CI 9.4 to 16.5). When positive scans were used to assist technologists, the MetaSystems platform had a sensitivity of 70.7% (94/133; 95% CI 62.2 to 78.3) and specificity of 89.0% (323/363; 95% CI 85.3 to 92.0). The manual microscopy method had a sensitivity of 79.7% (106/133; 95% CI 71.9 to 86.2) and specificity of 98.6% (358/363; 95% CI 96.8 to 99.6). The sensitivity of the MetaSystems platform was not impacted by smear grade or mycobacterial species. The majority (70.3%) of false positive smears had ≥2+ smear results with the MetaSystems platform. Further performance improvements are needed before the MetaSystems' automated fluorescent AFB slide reader can be used to assist microscopist in the clinical laboratory.

Association between body condition score fluctuations and pregnancy loss in Holstein cows

The objective of this study was to characterize the associations between body condition score (BCS) and BCS change (∆BCS), determined by an automated camera system during early lactation and close to artificial insemination (AI), and the subsequent pregnancy loss (PL) in Holstein cows. A secondary objective was to determine the impact of disease on PL, considering multiple time periods relative to AI. Data from 9,430 lactations in 6,884 Holstein cows in a commercial dairy operation located in Colorado, USA were included in this retrospective observational study. Cows were subject to first AI at about 80 DIM (primiparous) and 60 DIM (multiparous), following a double OvSynch protocol. Pregnancy diagnosis was performed via transrectal ultrasonography on day 32 ± 3 after AI and reconfirmed on day 80 ± 3 after AI. Cameras mounted on the sorting gate at each exit (n = 2) of the milking parlor generated BCS on a five-point scale with 0.1 increments. The BCS at calving (BCS1), 21 DIM (BCS21), 56 DIM (BCS56), AI resulting in pregnancy (BCSAI), and 90 d post AI (BCSAI90) were selected for the analyses and subsequently categorized as low (≤ lower quartile), moderate (interquartile range), and high (≥ upper quartile). Changes in BCS were calculated by periods of interest as change from calving to 21 DIM; change from calving to 56 DIM; change from 56 DIM to AI; and change from AI to 90 d post AI and assigned into categories to facilitate the analysis. Data were examined using logistic regression, considering parity category, season at calving and AI, DIM at AI, milk yield up to 60 DIM, and occurrence of disease as covariables. The logistic regression analyses indicated that the odds of PL were greater in cows in the low BCS category relative to cows in the high BCS category at 56 DIM (OR 95% CI = 1.41 [1.12-1.79]), AI (1.31 [1.05-1.65]), and 90 d post AI (1.38 [1.10-1.74]). Likewise, cows with large loss in BCS between calving and 21 DIM (1.46 [1.10-1.94]) and loss in BCS between AI and 90 d post AI (1.44 [1.15-1.81]) had greater odds of PL compared with cows with no loss of BCS within the same period. Occurrence of disease at all the time periods considered in the analysis had a consistent detrimental impact on maintenance of the pregnancy, supporting the concept that pre and postconceptional disease affects embryonic survival. Overall, low BCS, more pronounced reductions in BCS occurring closer to AI, and occurrence of disease resulted in greater PL in this Holstein population.

Performance of Contactless Respiratory Rate Monitoring by Albus HomeTM, an Automated System for Nocturnal Monitoring at Home: A Validation Study

Respiratory rate (RR) is a clinically important predictor of cardio-respiratory deteriorations. The mainstay of clinical measurement comprises the manual counting of chest movements, which is variable between clinicians and limited to sporadic readings. Emerging solutions are limited by poor adherence and acceptability or are not clinically validated. Albus Home^TM is a contactless and automated bedside system for nocturnal respiratory monitoring that overcomes these limitations. This study aimed to validate the accuracy of Albus Home compared to gold standards in real-world sleeping environments. Participants undertook overnight monitoring simultaneously using Albus Home and gold-standard polygraphy with thoraco-abdominal respiratory effort belts (SomnomedicsEU). Reference RR readings were obtained by clinician-count of polygraphy data. For both the Albus system and reference, RRs were measured in 30-s segments, reported as breaths/minute, and compared. Accuracy was defined as the percentage of RRs from the Albus system within ±2 breaths/minute of reference counts. Across a diverse validation set of 32 participants, the mean accuracy exceeded 98% and was maintained across different participant characteristics. In a Bland-Altman analysis, Albus RRs had strong agreement with reference mean differences and the limits of agreement of -0.4 and ±1.2 breaths/minute, respectively. Albus Home is a contactless yet accurate system for automated respiratory monitoring. Validated against gold -standard methods, it enables long-term, reliable nocturnal monitoring without patient burden.

The Use of Automated Machine Translation to Translate Figurative Language in a Clinical Setting: Analysis of a Convenience Sample of Patients Drawn From a Randomized Controlled Trial

BACKGROUND

Patients with limited English proficiency frequently receive substandard health care. Asynchronous telepsychiatry (ATP) has been established as a clinically valid method for psychiatric assessments. The addition of automated speech recognition (ASR) and automated machine translation (AMT) technologies to asynchronous telepsychiatry may be a viable artificial intelligence (AI)-language interpretation option.

OBJECTIVE

This project measures the frequency and accuracy of the translation of figurative language devices (FLDs) and patient word count per minute, in a subset of psychiatric interviews from a larger trial, as an approximation to patient speech complexity and quantity in clinical encounters that require interpretation.

METHODS

A total of 6 patients were selected from the original trial, where they had undergone 2 assessments, once by an English-speaking psychiatrist through a Spanish-speaking human interpreter and once in Spanish by a trained mental health interviewer-researcher with AI interpretation. 3 (50%) of the 6 selected patients were interviewed via videoconferencing because of the COVID-19 pandemic. Interview transcripts were created by automated speech recognition with manual corrections for transcriptional accuracy and assessment for translational accuracy of FLDs.

RESULTS

AI-interpreted interviews were found to have a significant increase in the use of FLDs and patient word count per minute. Both human and AI-interpreted FLDs were frequently translated inaccurately, however FLD translation may be more accurate on videoconferencing.

CONCLUSIONS

AI interpretation is currently not sufficiently accurate for use in clinical settings. However, this study suggests that alternatives to human interpretation are needed to circumvent modifications to patients' speech. While AI interpretation technologies are being further developed, using videoconferencing for human interpreting may be more accurate than in-person interpreting.

TRIAL REGISTRATION

ClinicalTrials.gov NCT03538860; https://clinicaltrials.gov/ct2/show/NCT03538860.

i-scope: A Compact automated fluorescence microscope for cell counting applications in low resource settings

Fluorescence microscopy has widespread applications across biological sciences. It has been routinely used for cell counting, which provides a preliminary diagnostic test for many infectious diseases. Conventional fluorescence microscopes are bulky, expensive, time-intensive and laborious. They often require trained operators to acquire and analyze data. We report a compact automated digital fluorescence microscopy system, i-scope, for cell counting applications. The i-scope employs a total internal reflection fluorescence (TIRF) mode of sample illumination, along with a brightfield mode. It has a magnification of 30X, an optical resolution of ~0.2 µm/pixel and offers sample scanning over 20 mm x 20 mm. A custom-written program enables automated image acquisition and analysis, thereby enhancing ease of operation. It has a compact form-factor and has been developed into a standalone system with a processing unit, screen, and other accessories to offer a portable and economic point-of-care diagnostic solution in low-resource settings. We analysed the performance of the i-scope for milk somatic cell enumeration and benchmarked it against that of a conventional fluorescence microscope.