Toward Data-Driven Radiation Oncology Using Standardized Terminology as a Starting Point: Cross-sectional Study

Background

The inability to seamlessly exchange information across radiation therapy ecosystems is a limiting factor in the pursuit of data-driven clinical practice. The implementation of semantic interoperability is a prerequisite for achieving the full capacity of the latest developments in personalized and precision medicine, such as mathematical modeling, advanced algorithmic information processing, and artificial intelligence approaches.

Objective

This study aims to evaluate the state of terminology resources (TRs) dedicated to radiation oncology as a prerequisite for an oncology semantic ecosystem. The goal of this cross-sectional analysis is to quantify the state of the art in radiation therapy specific terminology.

Methods

The Unified Medical Language System (UMLS) was searched for the following terms: radio oncology, radiation oncology, radiation therapy, and radiotherapy. We extracted 6509 unique concepts for further analysis. We conducted a quantitative analysis of available source vocabularies (SVs) and analyzed all UMLS SVs according to the route source, number, author, location of authors, license type, the lexical density of TR, and semantic types. Descriptive data are presented as numbers and percentages.

Results

The concepts were distributed across 35 SVs. The median number of unique concepts per SV was 5 (range 1-5479), with 14% (5/35) of SVs containing 94.59% (6157/6509) of the concepts. The SVs were created by 29 authors, predominantly legal entities registered in the United States (25/35, 71%), followed by international organizations (6/35, 17%), legal entities registered in Australia (2/35, 6%), and the Netherlands and the United Kingdom with 3% (1/35) of authors each. Of the total 35 SVs, 16 (46%) did not have any restrictions on use, whereas for 19 (54%) of SVs, some level of restriction was required. Overall, 57% (20/35) of SVs were updated within the last 5 years. All concepts found within radiation therapy SVs were labeled with one of the 29 semantic types represented within UMLS. After removing the stop words, the total number of words for all SVs together was 56,219, with a median of 25 unique words per SV (range 3-50,682). The total number of unique words in all SVs was 1048, with a median of 19 unique words per vocabulary (range 3-406). The lexical density for all concepts within all SVs was 0 (0.02 rounded to 2 decimals). Median lexical density per unique SV was 0.7 (range 0.0-1.0). There were no dedicated radiation therapy SVs.

Conclusions

We did not identify any dedicated TRs for radiation oncology. Current terminologies are not sufficient to cover the need of modern radiation oncology practice and research. To achieve a sufficient level of interoperability, of the creation of a new, standardized, universally accepted TR dedicated to modern radiation therapy is required.

Paradoxical positivity: Suicide notes use less distressed language than blogs about depression, suicidal thoughts, and even cooking

INTRODUCTION

While suicide attempts and deaths increase, research and assessment methods have stagnated in terms of increasing predictive power. Lexical analysis has been a useful method in descriptive suicide research, but may have utility for assessment and prediction.

OBJECTIVE

The present study used lexical analysis to examine language differences across a spectrum of psychological distress, with death by suicide as the most extreme indicator.

METHODS

Suicide writings were collected for 38 persons who died by suicide in two southern U.S. cities. Blog writings were collected from the "Depression" and "Suicide" categories of an online mental health discussion community and from food blogging communities (n = 38 each). Linguistic Inquiry and Word Count was used to calculate proportions of words that fell into lexical categories indicative of distress.

RESULTS

One-way ANOVA revealed significant differences in first-person singular pronouns and words related to positive emotions, negative emotions, religion, and death. Suicide notes were found to use less frequent lexical markers of distress than depression and suicide ideation blogs, and more frequent positive emotion language than all comparison groups.

CONCLUSION

The present findings suggest that deciding to take one's life may be associated with a decrease in distress, and possibly an increase in positive emotion, when compared to those experiencing depression and suicidal thoughts.

E-mental health in child psychiatry during COVID-19: an initial attitudinal study

OBJECTIVE

COVID-19 propelled e-mental health within the Australian health system. It is important to learn from this to inform mental healthcare during future crises.

METHOD

A lexical analysis was conducted of clinician reflections during COVID-19 as they delivered psychiatry services to children and families in New South Wales (n = 6) and transitioned to e-mental health.

RESULTS

E-mental health can extend the reach of, and access to psychiatry services, particularly for individuals disadvantaged by inequity. Yet e-mental health can be problematic. It is partly contingent on technological prowess, equipment, internet access as well as space and privacy. Relatedly, e-mental health can hinder clinician capacity to conduct examinations, monitor child development as well as assess risk and the need for child protection.

CONCLUSIONS

Given the benefits and limitations of e-mental health, a model that supports face-to-face mental healthcare and e-mental health may be of value. This model would require practical, yet flexible policies and protocols that protect the privacy of children and families, safeguard them from harm, and respect the needs and preferences of children, families and clinicians.

"More than words" - Longitudinal linguistic changes in the works of a writer diagnosed with semantic dementia

Leveraging recent advances in automated language analysis and anovel statistical approach utilizing an independent control group, we explored changes in lexical output across two published works of a man diagnosed with semantic dementia. We found significant increase in adverb usage and decline in familiarity, meaningfulness, age of acquisition and co-occurrence probability over 2 years. Collectively, these indices suggest that WR's narrative structure became progressively simpler, lexically less sophisticated, and that words commonly associated together no longer appeared in close proximity. Our study illustrates how degeneration of the semantic knowledge base impacts the production, content, and quality of literary works.

[The organization of medical service of the Interior Ministry of Russia in 2017]

The article presents a comprehensive analysis and assessment of the state of regional characteristics of morbidity with temporary disability, disability and mortality of employees of the internal affairs bodies of Russia. The source of the data was medical statistical reports generated by the medical and sanitary units of the Ministry of Internal Affairs of Russia using lexical analysis of medical documentation.

What kills us and what moves us: A comparative discourse analysis of heart disease and breast cancer

Introduction

Heart disease kills nearly 300,000 US women annually, while approximately 40,000 US women die of breast cancer. Breast cancer online patient communities are well known for their high engagement and emotional support. This exploratory study compared social media discourse on breast cancer with discourse related to heart disease.

Methods

Computer-assisted text analysis of two corpora composed of Twitter posts using #BreastCancer and #HeartDisease hashtags from December 2013 to December 2014. Lexical analysis (word and hashtag level) used AntConc software and lexicogrammatical analysis (style and stance) was conducted with DocuScope.

Results

The #BreastCancer corpus consisted of 592,046 posts, 57% of which were not original to the user (retweets). #HeartDisease had 269,769 posts (13% retweets). Social media discourse about #BreastCancer and #HeartDisease drew attention to women, new developments, appeals for help and disease risks. The #BreastCancer corpus incorporates gendered language and associations with art and activism, while posts about #HeartDisease were discussed scientifically in concert with other diseases. The #BreastCancer corpus uniquely included community-specific initialism hashtags. Stance analysis of the #BreastCancer corpus revealed more socially oriented posts, marked by language of constructive reasoning, inclusive language and abstract thought, while #HeartDisease corpus posts were more scholarly, used contingent and oppositional reasoning, language from institutional and academic registers, citations and meta-discourse.

Conclusion

The #HeartDisease social media community is less engaged, and content is less specific to both the disease and individual experience than #BreastCancer. Cultivating a women-focused heart disease online community might replicate some of the #BreastCancer community’s successes.

Question order and student understanding of structure and function

The relationship between structure and function is a core concept in physiology education. Written formative assessments can provide insight into student learning of the structure and function relationship, which can then inform pedagogy. However, question order may influence student explanations. We explored how the order of questions from different cognitive levels affects student explanations. A junior level General Physiology class was randomly split in half. One-half of the students answered, "Define the principle: form reflects function," followed by "Give an example of the principle: form reflects function" (format DX), whereas the other half answered, "Give an example of the principle: form reflects function," followed by "Define the principle: form reflects function" (format XD). Human grading and computerized lexical analysis were used to evaluate student responses. Two percent of students in the format DX group related structure and function in their definition, whereas 48% of students related structure and function in their examples. In the format XD group, 17% related structure and function in their definition, and 26% related structure and function in their example of the principle. Overall, students performed better on the last question in the sequence, which may be evidence for conceptual priming. Computerized lexical analysis revealed that students draw on only a few levels of organization and may be used by instructors to quickly assess the levels of organization students use in their responses. Written assessment coupled with lexical analysis has the potential to reveal student understanding of core concepts in anatomy and physiology education.

Self-reference in psychosis and depression: a language marker of illness

BACKGROUND

Language use is of increasing interest in the study of mental illness. Analytical approaches range from phenomenological and qualitative to formal computational quantitative methods. Practically, the approach may have utility in predicting clinical outcomes. We harnessed a real-world sample (blog entries) from groups with psychosis, strong beliefs, odd beliefs, illness, mental illness and/or social isolation to validate and extend laboratory findings about lexical differences between psychosis and control subjects.

METHOD

We describe the results of two experiments using Linguistic Inquiry and Word Count software to assess word category frequencies. In experiment 1, we compared word use in psychosis and control subjects in the laboratory (23 per group), and related results to subject symptoms. In experiment 2, we examined lexical patterns in blog entries written by people with psychosis and eight comparison groups. In addition to between-group comparisons, we used factor analysis followed by clustering to discern the contributions of strong belief, odd belief and illness identity to lexical patterns.

RESULTS

Consistent with others' work, we found that first-person pronouns, biological process words and negative emotion words were more frequent in psychosis language. We tested lexical differences between bloggers with psychosis and multiple relevant comparison groups. Clustering analysis revealed that word use frequencies did not group individuals with strong or odd beliefs, but instead grouped individuals with any illness (mental or physical).

CONCLUSIONS

Pairing of laboratory and real-world samples reveals that lexical markers previously identified as specific language changes in depression and psychosis are probably markers of illness in general.

Combining Video, Audio and Lexical Indicators of Affect in Spontaneous Conversation via Particle Filtering

We present experiments on fusing facial video, audio and lexical indicators for affect estimation during dyadic conversations. We use temporal statistics of texture descriptors extracted from facial video, a combination of various acoustic features, and lexical features to create regression based affect estimators for each modality. The single modality regressors are then combined using particle filtering, by treating these independent regression outputs as measurements of the affect states in a Bayesian filtering framework, where previous observations provide prediction about the current state by means of learned affect dynamics. Tested on the Audio-visual Emotion Recognition Challenge dataset, our single modality estimators achieve substantially higher scores than the official baseline method for every dimension of affect. Our filtering-based multi-modality fusion achieves correlation performance of 0.344 (baseline: 0.136) and 0.280 (baseline: 0.096) for the fully continuous and word level sub challenges, respectively.