Automatic selection of spoken language biomarkers for dementia detection.

This paper analyzes diverse features extracted from spoken language to select the most discriminative ones for dementia detection. We present a two-step feature selection (FS) approach: Step 1 utilizes filter methods to pre-screen features, and Step 2 uses a novel feature ranking (FR) method, referred to as dual dropout ranking (DDR), to rank the screened features and select spoken language biomarkers. The proposed DDR is based on a dual-net architecture that separates FS and dementia detection into two neural networks (namely, the operator and selector). The operator is trained on features obtained from the selector to reduce classification or regression loss. The selector is optimized to predict the operator's performance based on automatic regularization. Results show that the approach significantly reduces feature dimensionality while identifying small feature subsets that achieve comparable or superior performance compared with the full, default feature set. The Python codes are available at https://github.com/kexquan/dual-dropout-ranking.

Intelligent surgical workflow recognition for endoscopic submucosal dissection with real-time animal study.

Recent advancements in artificial intelligence have witnessed human-level performance; however, AI-enabled cognitive assistance for therapeutic procedures has not been fully explored nor pre-clinically validated. Here we propose AI-Endo, an intelligent surgical workflow recognition suit, for endoscopic submucosal dissection (ESD). Our AI-Endo is trained on high-quality ESD cases from an expert endoscopist, covering a decade time expansion and consisting of 201,026 labeled frames. The learned model demonstrates outstanding performance on validation data, including cases from relatively junior endoscopists with various skill levels, procedures conducted with different endoscopy systems and therapeutic skills, and cohorts from international multi-centers. Furthermore, we integrate our AI-Endo with the Olympus endoscopic system and validate the AI-enabled cognitive assistance system with animal studies in live ESD training sessions. Dedicated data analysis from surgical phase recognition results is summarized in an automatically generated report for skill assessment.

Speaker-turn aware diarization for speech-based cognitive assessments.

Introduction: Speaker diarization is an essential preprocessing step for diagnosing cognitive impairments from speech-based Montreal cognitive assessments (MoCA). Methods: This paper proposes three enhancements to the conventional speaker diarization methods for such assessments. The enhancements tackle the challenges of diarizing MoCA recordings on two fronts. First, multi-scale channel interdependence speaker embedding is used as the front-end speaker representation for overcoming the acoustic mismatch caused by far-field microphones. Specifically, a squeeze-and-excitation (SE) unit and channel-dependent attention are added to Res2Net blocks for multi-scale feature aggregation. Second, a sequence comparison approach with a holistic view of the whole conversation is applied to measure the similarity of short speech segments in the conversation, which results in a speaker-turn aware scoring matrix for the subsequent clustering step. Third, to further enhance the diarization performance, we propose incorporating a pairwise similarity measure so that the speaker-turn aware scoring matrix contains both local and global information across the segments. Results: Evaluations on an interactive MoCA dataset show that the proposed enhancements lead to a diarization system that outperforms the conventional x-vector/PLDA systems under language-, age-, and microphone-mismatch scenarios. Discussion: The results also show that the proposed enhancements can help hypothesize the speaker-turn timestamps, making the diarization method amendable to datasets without timestamp information.

The Hong Kong Grocery Shopping Dialog Task (HK-GSDT): A Quick Screening Test for Neurocognitive Disorders.

The Hong Kong Grocery Shopping Dialog Task (HK-GSDT) is a short and easy-to-administer cognitive test developed for quickly screening neurocognitive disorders (NCDs). In the test, participants are instructed to do a hypothetical instrumental activity of daily living task of purchasing ingredients for a dish from a grocery store and verbally describe the specific shopping procedures. The current study aimed to validate the test with a sample of 545 Hong Kong older adults (58.8% female; aged 73.4 ± 8.37 years), including 464 adults with normal cognitive function, 39 with mild NCD, and 42 with major NCD. Demographic characteristics (i.e., sex, age, education) and clinical diagnosis of cognitive states (i.e., major NCD, mild NCD, and normal aging) were collected. Cognitive functioning was measured using the HK-GSDT and several standardized NCD-screening tests. The results showed good reliability (i.e., internal consistency) and structural validity in the HK-GSDT. It discriminated among different cognitive conditions, particularly between major NCDs and the other conditions, as effectively as did the existing standardized neurocognitive tests (e.g., Montreal Cognitive Assessment, Hong Kong List Learning Test). Moreover, the HK-GSDT explained additional variance of cognitive condition on top of those standardized neurocognitive tests. These results indicate that the HK-GSDT can be used alone, or in combination with other tests, to screen for NCDs.

An Integrated Approach of Machine Learning and Systems Thinking for Waiting Time Prediction in an Emergency Department.

OBJECTIVE: The objective of this study is to apply machine learning algorithms for real-time and personalized waiting time prediction in emergency departments. We also aim to introduce the concept of systems thinking to enhance the performance of the prediction models. METHODS: Four popular algorithms were applied: (i) stepwise multiple linear regression; (ii) artificial neural networks; (iii) support vector machines; and (iv) gradient boosting machines. A linear regression model served as a baseline model for comparison. We conducted computational experiments based on a dataset collected from an emergency department in Hong Kong. Model diagnostics were performed, and the results were cross-validated. RESULTS: All the four machine learning algorithms with the use of systems knowledge outperformed the baseline model. The stepwise multiple linear regression reduced the mean-square error by almost 15%. The other three algorithms had similar performances, reducing the mean-square error by approximately 20%. Reductions of 17 - 22% in mean-square error due to the utilization of systems knowledge were observed. DISCUSSION: The multi-dimensional stochasticity arising from the ED environment imposes a great challenge on waiting time prediction. The introduction of the concept of systems thinking led to significant enhancements of the models, suggesting that interdisciplinary efforts could potentially improve prediction performance. CONCLUSION: Machine learning algorithms with the utilization of the systems knowledge could significantly improve the performance of waiting time prediction. Waiting time prediction for less urgent patients is more challenging.

The molecular tweezer CLR01 inhibits aberrant superoxide dismutase 1 (SOD1) self-assembly in vitro and in the G93A-SOD1 mouse model of ALS.

Mutations in superoxide dismutase 1 (SOD1) cause 15-20% of familial amyotrophic lateral sclerosis (fALS) cases. The resulting amino acid substitutions destabilize SOD1's protein structure, leading to its self-assembly into neurotoxic oligomers and aggregates, a process hypothesized to cause the characteristic motor-neuron degeneration in affected individuals. Currently, effective disease-modifying therapy is not available for ALS. Molecular tweezers prevent formation of toxic protein assemblies, yet their protective action has not been tested previously on SOD1 or in the context of ALS. Here, we tested the molecular tweezer CLR01-a broad-spectrum inhibitor of the self-assembly and toxicity of amyloid proteins-as a potential therapeutic agent for ALS. Using recombinant WT and mutant SOD1, we found that CLR01 inhibited the aggregation of all tested SOD1 forms in vitro Next, we examined whether CLR01 could prevent the formation of misfolded SOD1 in the G93A-SOD1 mouse model of ALS and whether such inhibition would have a beneficial therapeutic effect. CLR01 treatment decreased misfolded SOD1 in the spinal cord significantly. However, these histological findings did not correlate with improvement of the disease phenotype. A small, dose-dependent decrease in disease duration was found in CLR01-treated mice, relative to vehicle-treated animals, yet motor function did not improve in any of the treatment groups. These results demonstrate that CLR01 can inhibit SOD1 misfolding and aggregation both in vitro and in vivo, but raise the question whether such inhibition is sufficient for achieving a therapeutic effect. Additional studies in other less aggressive ALS models may be needed to determine the therapeutic potential of this approach.

How can we better use Twitter to find a person who got lost due to dementia?

Twitter is a social media platform for online message sharing. The aim of this study is to evaluate the effectiveness of using Twitter to search for people who got lost due to dementia. The online messages on Twitter, i.e., tweets, were collected through an Application Programming Interface. Contents of the tweets were analysed. The personal characteristics, features of tweets and types of Twitter users were collected to investigate their associations with whether a person can be found within a month. Logistic regression was used to identify the features that were useful in finding the missing people. Results showed that the young age of the persons with dementia who got lost, having tweets posted by police departments, and having tweets with photos can increase the chance of being found. Social media is reshaping the human communication pathway, which may lead to future needs on a new patient-care model.

Cardiovascular mortality in hypertensive patients newly prescribed perindopril vs. lisinopril: a 5-year cohort study of 15,622 Chinese subjects.

BACKGROUND: Perindopril and lisinopril are two common ACE inhibitors prescribed for management of hypertension. Few studies have evaluated their comparative effectiveness to reduce mortality. This study compared the all-cause and cardiovascular related mortality among patients newly prescribed ACE inhibitors. METHODS: All adult patients newly prescribed perindopril or lisinopril from 2001 to 2005 in all public clinics or hospitals in Hong Kong were retrospectively evaluated, and followed up until 2010. Patients prescribed the ACE inhibitors for less than a month were excluded. The all-cause mortality and cardiovascular-specific (i.e. coronary heart disease, heart failure and stroke) mortality were compared. Cox proportional hazard regression model was used to assess the mortality, controlling for age, sex, socioeconomic status, patient types, the presence of comorbidities, and medication adherence as measured by the proportion of days covered. An additional model using propensity scores was performed to minimize indication bias. RESULTS: A total of 15,622 patients were included in this study, in which 6910 were perindopril users and 8712 lisinopril users. The all-cause mortality (22.2% vs. 20.0%, p<0.005) and cardiovascular mortality (6.5% vs. 5.6%, p<0.005) were higher among lisinopril users than perindopril users. From regression analyses, lisinopril users were 1.09-fold (95% C.I. 1.01-1.16) and 1.18-fold (95% C.I. 1.02-1.35) more likely to die from any-cause and cardiovascular diseases, respectively. Age-stratified analysis showed that this significant difference was observed only among patients aged >70 years. The additional models controlled for propensity scores yielded comparable results. CONCLUSIONS: The long-term all-cause and cardiovascular related mortality rates of lisinopril users was significantly different from those of perindopril users. These findings showed that intra-class variation on mortality exists among ACE inhibitors among those aged 70 years or older. Future studies should consider a longer, large-scale randomized controlled trial to compare the effectiveness between different medications in the ACEI class, especially among the elderly.