Correlation between Task Complexity and Accuracy of Cortical Language Mapping

Introduction

This study aimed to investigate whether task complexity affects the accuracy of mapping language functions during direct cortical stimulation mapping (DCS). The researchers hypothesized that due to the reduced computational ability of neurons in the cortex infiltrated by brain tumors, complex language tasks (such as naming multisyllabic words) may increase the error rate.

Paper Source

This study was conducted by Alexa Semonche and colleagues from the Department of Neurological Surgery at the University of California, San Francisco (UCSF). The results were published online on May 7, 2024, in the journal Neurosurgery. This paper was previously presented as an abstract and electronic poster at the American Association of Neurological Surgeons (AANS) annual meeting on April 21, 2023.

Research Methods

Study Subjects

The researchers retrospectively analyzed data from 74 patients who underwent resection of unilateral language-dominant hemispheric gliomas (WHO grade II-IV) and underwent awake language mapping at UCSF between 2017 and 2021.

Research Procedure

1. Preoperatively assessed patients’ language function and recorded their performance on picture naming (PN) and word reading (WR) tasks.
2. During surgery, performed DCS language mapping and recorded the accuracy rates for PN and WR tasks.
3. Divided PN and WR trials into three groups based on the number of syllables: 1 syllable, 2 syllables, and 3+ syllables.
4. Analyzed the error rates for each group to explore the correlation between syllable count and error rate.
5. Applied multivariate logistic regression to analyze factors related to error rate.
6. Analyzed the relationship between error rate and extent of resection.

Data Analysis

The researchers used Python and R software for data analysis, employing statistical methods such as chi-square tests and multivariate logistic regression.

Main Results

1. A total of 2,643 PN trials and 978 WR trials were analyzed.
2. The overall error rate for PN was 5.2%, and for WR, it was 6.0%.
3. An increase in the number of syllables was significantly associated with a higher PN error rate (p=0.001) but not with the WR error rate (p=0.807).
4. Multivariate analysis showed that for every additional syllable, the odds ratio for the PN error rate increased by 2.4 times (p<0.001). For every one-year increase in age, the odds ratio for the error rate increased by 1.013 times (p=0.043).
5. The severity of preoperative language deficits was positively correlated with intraoperative multisyllabic PN error rate (p<0.001, r2=0.1776).
6. There was no significant correlation between error rate and extent of resection (p=0.949).

Research Significance

1. Discovered that complex (multisyllabic) language tasks may exceed the computational ability of neurons in the cortex infiltrated by gliomas.
2. Suggested modifications to the DCS language mapping paradigm, grouping stimulation sites based on task complexity to more accurately assess functional areas.
3. Expanded the field of cancer neuroscience by revealing how gliomas alter the cortical processing of language.
4. The study results provide clinical guidance for optimizing surgical strategies and evaluating postoperative language function recovery in patients.

Research Highlights

1. Large-sample, single-center study providing a detailed analysis of DCS language mapping.
2. Innovatively associated task complexity (syllable count) with mapping accuracy.
3. Proposed modifications to the DCS language mapping strategy to promote precise preservation of language areas.
4. Expanded the field of “cancer neuroscience” by exploring how gliomas affect cortical language function.

This study offers valuable suggestions for improving the DCS language mapping paradigm and provides new perspectives for neurosurgery and neurorehabilitation. Future research should further explore the molecular biological mechanisms underlying the reduced computational ability of neurons in glioma-infiltrated cortical regions.