In Brief

We propose a novel multimodal task for cortico-subcortical cognitive mapping, designed to be maximally sensitive by simultaneously covering various language and cognitive pathways. The task presents four pictures (typically objects) in the corners of a slide, which relate to a subject or verb (or both) displayed in the center text box to form a sentence. This approach tests object naming, semantics, reading and/or hearing, and syntax simultaneously, thereby reducing the probability of false negatives.

Introduction

In awake language mapping for resection of supratentorial lesions in eloquent areas, object naming has long been the gold standard task during cortical and subcortical stimulation. However, using only this task can lead to false negatives, increasing the risk of damaging eloquent language areas. It is insufficient to detect errors in different language pathways, such as semantics, phonology, and syntax, especially subcortically, where various language fascicles overlap and intersect. In these regions, multiple language functions must be monitored simultaneously.

Most tasks analyze only one input mode:

  • One visual input for object naming, action naming, semantic odd picture out, and semantic association.

  • One auditory input for repetition and diadochokinesis.

  • One reading input for phonological and semantic odd word out, sentence completion, and semantic association.

In daily life, visual, auditory, and possibly reading inputs usually occur together. Additionally, verbal and motor outputs often happen simultaneously, such as speaking with gestures or indications. Most tasks use different input and output modes separately to evaluate specificity, often at the expense of sensitivity. Logically, highly sensitive tasks should precede the search for specificity.

Furthermore, simultaneous multimodal input saves time, which is crucial during tasks with the standard 4-second stimulation timeframe. Hence, there is an obvious need for a novel task that evaluates different cognitive pathways, including their input modes, providing broad scanning and high sensitivity.

Background

Since our team started in 2002 with awake language mapping , it seemed prompltly obvious that Picture Naming was insufficient to map cortical and subcortical structures in a reliable way: postoperatively marked deficiency in semantics was observed in patients where Object Naming remained correct during the whole testing. So in a first stage, we moved to the incorporation of Semantic Odd Picture Out tasks or equivalent Pyramid and Palm Tree test, yielding an evaluation of the visual semantic pathways. A first further step was to present 4 items( 1 in each corner of the screen) instead of the classical 3 objects : this yielded simultaneous evaluation of the visual fields (3).

The need of combining different tasks, analysing phonological, semantic, syntactic and articulatory pathways became obvious: this was in 2012 the craddle of the DuLIP ( Dutch Linguistic Intraoperative Protocol), elaborated with Rotterdam and Brussels, and finalised in 2014 after full standardisation * .

A next step in 2015 was the proposal of a multimodal testing with different DuLIP tasks sequencing (“ Quick Mixed Test”), with alternation of e.g. a repetion task, Object naming, calculation and Line Bisection; however, this sequenced structure of tasks appeared us to be suboptimal, since during the evaluation of one language pathway, another could be neglected. Moreover, this presentation mode is more tiring fort he patient.

Different tasks succeeding each other are time consuming, masking evaluation of other functions when examing one specific function, boring for the patient and susceptible to subjectivity of the evaluator; moreover, a choice has to be made which task where, with possibility of selection error.

Since the choice had to be made between phonological, semantic or syntactic tasks, the idea arose of the combined monitoring of different pathways in one task, presenting a written subject in the centre of the slide, which has to be associated with one of the 4 objects, and to form a short sentence ( with addition of verb generation, and obtention of a syntactic structure).

Later written items could be replaced or completed by spoken items; Objects, Subjects and Verbs can be interchanged : this offers a multiplicity of possible combinations, in which the most relevant has to be selected.