Alternate fluency in Parkinson’s disease: A machine learning analysis

Can total electrical energy (TEED) after subthalamic DBS alter verbal fluency in Parkinson's disease patients? A preliminary evidence

OBJECTIVE

Deep brain stimulation (DBS) of the subthalamic nucleus (STN) improves motor outcomes in Parkinson's disease (PD) but may have adverse long-term effects on specific cognitive domains. The aim of this study was to investigate the association between total electrical energy (TEED) delivered by DBS and postoperative changes in verbal fluency.

METHODS

Seventeen PD patients undergoing bilateral STN-DBS were assessed with the Alternate Verbal Fluency Battery (AVFB), which includes phonemic (PVF), semantic (SVF), and alternate verbal fluency (AVF) tests, before surgery (T0) and after 6 (T1) and 12 months (T2). Bilateral TEED and average TEEDM were recorded at T1 and T2. For each AVFB measurement, changes from T0 to T1 (Δ-01) and from T0 to T2 (Δ-02) were calculated.

RESULTS

At T1, PVF (p = 0.007) and SVF scores (p = 0.003) decreased significantly. TEED measures at T1 and T2 were unrelated to Δ-01 and Δ-02 scores, respectively. However, an inverse, marginally significant association was detected between the TEEDM and Δ-01 scores for the AVF (p = 0.041, against an αadjusted = 0.025).

CONCLUSIONS

In conclusion, the present reports provide preliminary evidence that TEED may not be responsible or only slightly responsible for the decline in VF performance after STN-DBS in PD.

Psychometrics and diagnostics of the Italian version of the Alternate Verbal Fluency Battery (AVFB) in non-demented Parkinson's disease patients

BACKGROUND

Verbal fluency (VF) tasks are known as suitable for detecting cognitive impairment (CI) in Parkinson's disease (PD). This study thus aimed to evaluate the psychometrics and diagnostics of the Alternate Verbal Fluency Battery (AVFB) by Costa et al. (2014) in an Italian cohort of non-demented PD patients, as well as to derive disease-specific cut-offs for it.

METHODS

N = 192 non-demented PD patients were screened with the Montreal Cognitive Assessment (MoCA) and underwent the AVFB-which includes phonemic, semantic and alternate VF tests (PVF; SVF; AVF), as well as a Composite Shifting Index (CSI) reflecting the "cost" of shifting from a single- to a double-cued VF task. Construct validity and diagnostics were assessed for each AVFB measure against the MoCA. Internal reliability and factorial validity were also tested.

RESULTS

The MoCA proved to be strongly associated with PVF, SVF and AVF scores, whilst moderately with the CSI. The AVFB was internally consistent and underpinned by a single component; however, an improvement in both internal reliability and fit to its factorial structure was observed when dropping the CSI. Demographically adjusted scores on PVF, SVF and AVF tests were diagnostically sound in detecting MoCA-defined cognitive impairment, whilst this was not true for the CSI. Disease-specific cut-offs for PVF, SVF and AVF tests were derived.

DISCUSSION

In conclusion, PVF, SVF and AVF tests are reliable, valid and diagnostically sound instruments to detect cognitive impairment in non-demented PD patients and are therefore recommended for use in clinical practice and research.

Human-like problem-solving abilities in large language models using ChatGPT

Backgrounds

The field of Artificial Intelligence (AI) has seen a major shift in recent years due to the development of new Machine Learning (ML) models such as Generative Pre-trained Transformer (GPT). GPT has achieved previously unheard-of levels of accuracy in most computerized language processing tasks and their chat-based variations.

Aim

The aim of this study was to investigate the problem-solving abilities of ChatGPT using two sets of verbal insight problems, with a known performance level established by a sample of human participants.

Materials and methods

A total of 30 problems labeled as "practice problems" and "transfer problems" were administered to ChatGPT. ChatGPT's answers received a score of "0" for each incorrectly answered problem and a score of "1" for each correct response. The highest possible score for both the practice and transfer problems was 15 out of 15. The solution rate for each problem (based on a sample of 20 subjects) was used to assess and compare the performance of ChatGPT with that of human subjects.

Results

The study highlighted that ChatGPT can be trained in out-of-the-box thinking and demonstrated potential in solving verbal insight problems. The global performance of ChatGPT equalled the most probable outcome for the human sample in both practice problems and transfer problems as well as upon their combination. Additionally, ChatGPT answer combinations were among the 5% of most probable outcomes for the human sample both when considering practice problems and pooled problem sets. These findings demonstrate that ChatGPT performance on both set of problems was in line with the mean rate of success of human subjects, indicating that it performed reasonably well.

Conclusions

The use of transformer architecture and self-attention in ChatGPT may have helped to prioritize inputs while predicting, contributing to its potential in verbal insight problem-solving. ChatGPT has shown potential in solving insight problems, thus highlighting the importance of incorporating AI into psychological research. However, it is acknowledged that there are still open challenges. Indeed, further research is required to fully understand AI's capabilities and limitations in verbal problem-solving.