A "Verbal Thermometer" for Assessing Neurodegenerative Disease: Automated Measurement of Pronoun and Verb Ratio from Speech

Abstract

Clinicians often use speech to characterize neurodegenerative disorders. Such characterizations require clinical judgment, which is subjective and can require extensive training. Quantitative Production Analysis (QPA) can be used to obtain objective quantifiable assessments of patient functioning. However, such human-based analyses of speech are costly and time consuming. Inexpensive off-the-shelf technologies such as speech recognition and part of speech taggers may avoid these problems. This study evaluates the ability of an automatic speech to text transcription system and a part of speech tagger to assist with measuring pronoun and verb ratios, measures based on QPA. Five participant groups provided spontaneous speech samples. One group consisted of healthy controls, while the remaining groups represented four subtypes of frontotemporal dementia. Findings indicated measurement of pronoun and verb ratio was robust despite errors introduced by automatic transcription and the tagger and despite these off-the-shelf products not having been trained on the language obtained from speech of the included population. Clinical relevance Linguistic differences in individuals with neurodegenerative disease can be subtle. Automated measurement of pronoun and verb ratio (PR and VR) showed promise despite the early stage of pipeline development using off-theshelf tools. Intraclass correlations between automatic versus human measures of PR and VR were in the moderate to excellent range as measured by intra-class correlation.

Primary Progressive Aphasia: Toward a Pathophysiological Synthesis

“…Identification and stratification into clinical trials at the very earliest stage of disease will be vital if treatments are to be effective. The use of ecologically relevant measures that quantify aspects of speech output have garnered recent interest in terms of both diagnosis and disease tracking from early stages [30][31][32][33], leading to the suggestion that automated analysis of speech might be used as a 'verbal thermometer' for PPA and FTD [34]. There is also considerable interest in the use of 'wet biomarkers' in measuring disease burden and intensity.…”

Description of connected speech across different elicitation tasks in the logopenic variant of primary progressive aphasia

“…talkbank.org/discourse/). For instance, a recent study by Jarrold et al (2020) has shown that automated analysis could robustly measure pronoun and verb ratios in neurodegenerative populations, two measures present in the QPA.…”

Symptom-led staging for primary progressive aphasia

“…A staging system for PPA could provide a bridge to the development of bespoke functional scales of daily life communication 77 that could in turn serve as outcome measures for clinical trials in these syndromes and language-led AD clinical phenotypes 78 . Looking forward, we envisage that particular symptoms identified as harbingers of early-stage PPA might inform the development of diagnostic 'cognitive stress tests' based on the analysis of patients' spontaneous conversation [79][80][81] or requiring production and/or understanding of speech in multi-talker environments or other demanding conditions 52,82 . However, the overriding value of a PPA staging system, as attested to in the qualitative analysis, will lie in the personalised care of individual patients and families navigating these illnesses -to help them signpost their journey, harness all appropriate supports and treatments, and plan for the future.…”