Brain, Vol. 125, No. 11, 2537-2548,
November 2002
© 2002 Oxford University Press
Hierarchical versus parallel processing in tactile object recognition
A behavioural–neuroanatomical study of aperceptive tactile agnosia
1 Department of Neurology and 2 Institute of Radiology, Kantonsspital St Gallen, CH-9007 St Gallen, SwitzerlandCorrespondence to: Bruno Weder, Klinik für Neurologie, Kantonsspital St Gallen, CH-9007 St Gallen, Switzerland E-mail: bruno.weder@kssg.ch
Received December 19, 2001. Revised June 3, 2002. Accepted June 6, 2002.
 |
Summary |
|---|
 Top Summary IntroductionMaterial and methodsResultsDiscussionReferences |
|---|
The organization of the normal perceptual processing subserving tactile object recognition is poorly understood. While perceptual deficits associated with cases of tactile agnosia may pinpoint sites of critical interference with normal tactile information processing, the precise character of such deficits remains unclear. The aim of the present study was to explore the behavioural and neuroanatomical correlates of perceptual disturbances in two cases of unilateral aperceptive tactile agnosia. Perception of microgeometrical and macrogeometrical features was tested using an alternative forced choice paradigm. While both patients were impaired in the assessment of microgeometrical properties of objects (i.e. detecting subtle differences in grating profiles), one patient showed an additional deficit in the perception of macrogeometrical properties of objects (i.e. detecting differences in length of cuboids). The pattern of perceptual deficits for both patients suggested a severely compromised (if not totally lost) ability to recognize everyday objects. Perceptual performance improved when the patients had complementary tactile information (i.e. for intramodal comparison), despite a persistent inability to explicitly name the objects. That is, the patients were able to recognize objects, but only implicitly. Improved perceptual performance was also observed when complementary visual information was available (i.e. transmodal information transfer). In this case, the perceptual improvement was accompanied by a corresponding improvement in explicit object recognition. High resolution MRIs identified lesions in the postcentral gyrus in both patients, and additionally in the secondary somatosensory area (SII) and the posterior parietal cortex in the more severely affected patient. The results demonstrate that the underlying failure in tactile agnosia is mainly impaired perception of microgeometrical properties of objects due to a lesion of primary sensory cortex. The related neuroanatomical findings suggest a degradation of serial information processing within postcentral gyrus. In one case tactile agnosia was almost complete due to additionally impaired perception of macrogeometrical properties of objects, which correlated with the extension of lesion to the posterior parietal cortex. Importantly, the findings indicate traces of two distributed networks for tactile information processing and the associated parallel processing of complementary micro- and macrogeometrical information within postcentral gyrus and posterior parietal lobe.
Keywords: high resolution MRI; macrogeometry; microgeometry; parietal cortex; tactile agnosia
Abbreviations: BA = Brodmann area; CBA = computerized brain atlas; fMRI = functional MRI; OR = odds ratio; PC = probability of correct answer
|
Introduction |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Tactile object recognition can be viewed as complex information processing that evolves from modality-specific perception of different classes of object features into recognition of objects explored. In addition, knowledge about these objects, based on a supra- or amodal memory store (Mesulam, 1998
), is thought to contribute to final object recognition. Little is known about the behavioural and neuroanatomical basis of these perceptual processes in normal tactile object recognition.
While perceptual deficits associated with cases of tactile agnosia may pinpoint sites of critical interference with normal tactile information processing, the precise character of such deficits remains unclear. Indeed, a disturbance in tactile recognition has been designated by different terms, i.e. astereognosis (Roland, 1976) and tactile agnosia (Caselli, 1997). This lack of uniformity in diagnostic designations merely reflects disagreement concerning the relative contributions made to the disturbance by patho-physiological mechanisms on the one hand, or by levels of information processing on the other. In as much as the terminology describing such a disturbance has been used interchangeably, it is evident that there is little clarity concerning the precise perceptual deficits associated with a disturbance in tactile object recognition. Although disturbed tactile object recognition occurs only rarely in the absence of elementary somaesthetic dysfunction (Mauguiere and Isnard, 1995), an impairment in the perception of elementary sensations (such as light touch or position sense) is unlikely to account fully for the disorder (Wiebers et al., 1998). Thus, we apply the term to describe a disturbed processing of sensory information beyond an elementary sensory failure. Furthermore, a differentiation of aperceptive from associative tactile agnosia has been taken into consideration, as recently suggested (Mauguiere and Isnard, 1995; Platz, 1996; Reed et al., 1996). Accordingly, aperceptive tactile agnosia results from a faulty perceptual synthesis of micro- and macrogeometrical features of objects, whereas associative tactile agnosia arises when intact perceptual extraction of object features fails to associate with semantic knowledge of objects (Damasio, 1992).
Thus, successful tactile object recognition seems to depend critically on parallel information processing of micro- and macrogeometrical properties. Such properties correspond to the length of axes and surface qualities of real world objects, respectively (Morley et al., 1983; Roland and Mortensen, 1987). Earlier lesion studies (Roland, 1987) and more recent functional imaging studies (Seitz et al., 1991; O’Sullivan et al., 1994; Ledberg et al., 1995; Roland et al., 1998; Binkofski et al., 1999b) indicate separate somatosensory association areas involved in the perception of microgeometrical and macrogeometrical properties. The superior parietal cortex has been shown to be predominantly activated during length and shape discrimination, while the parietal operculum has been shown to be involved during roughness discrimination. Correlative clinical and neuroimaging studies demonstrated causative lesions of both the primary somatosensory cortex (Roland, 1976) and posterior parts of the parietal cortex (Pause et al., 1989; Bassetti et al., 1993; Binkofski et al., 2001). Cases assessed as associative tactile agnosia were variably associated with lesions in the inferior parietal cortex (Reed et al., 1996) or primary somatosensory and superior parietal cortex (Platz, 1996), as detected by MRI.
Here we report the results of a behavioural–neuroanatomical investigation of two cases of unilateral tactile agnosia. The specific aims of the study were: first, to examine the differential contributions made to tactile object recognition by microgeometrical and macrogeometrical object properties; secondly, to assess the effects on tactile object recognition of intramodal comparison and crossmodal information transfer; and thirdly, to examine the relation between the extent of disturbed somaesthetic function and the extent of parietal lesions.
|
Material and methods |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Subjects
Patients
The two patients presented unilateral tactile agnosia on the right side. One showed a discrete spastic hand paresis. Handedness was assessed by the Edinburgh Handedness Inventory (Oldfield, 1971
). Patient 1 was a 67-year-old, right-handed woman, with the sequelae of a left parietal ischaemic infarction (which occurred 30 years ago). She had fully recovered from the infarction with the exception of a persistent inability to recognize everyday objects by tactile exploration with the right hand. MRIs disclosed an isolated, large defect within the postcentral gyrus, extending from the Sylvian to the longitudinal fissure. In particular, the lesion included the supramarginal gyrus and the superior parietal lobule (also see Results). Neurological examination was normal apart from disturbed two-point discrimination and tactile agnosia of the right hand (Table 1). Neuropsychological evaluation revealed moderately disturbed executive functions, while cognitive functions, notably memory and attention, were intact.
|
Patient 2 was a 29-year-old, right-handed man who had experienced recurrent focal motor seizures of the right upper limb for 8 years. Manifestation consisted of unpleasant sensations of the right hand and temporary weakness of the whole right upper extremity. Cranial computed tomography (CT) revealed an arteriovenous malformation in the left temporo-parietal lobe. Initially, the patient refused treatment, but, following a further sensorimotor seizure with secondary generalization 5 years later, he was treated with antiepileptic medication, i.e. carbamazepine. Thereafter, he received several embolizations of the arteriovenous malformation by endovascular techniques leaving neurological sequelae as noted below and most likely of ischaemic origin (Qureshi et al., 2000
). The data presented below were obtained following a stable course of 2 years since then. A control MRI showed a partially embolized lesion in the left postcentral gyrus extending subcortically to the border of lateral ventricle and ventrally to precentral gyrus. In contrast to Patient 1, the supramarginal gyrus and the superior parietal lobule were spared in this patient (see Results). He showed slight spasticity of the right upper extremity with exaggerated tendon reflexes and weakened closure of the fist. The main finding was a marked tactile agnosia of the right hand. Elementary sensations were not affected and two-point discrimination was moderately disturbed (Table 1). The results of neuropsychological testing were unremarkable and within the normal range, especially as regards selective attention, cognitive flexibility and memory.
Controls
Data from control subjects served as normative data for the length and roughness discrimination tasks. For the former task, controls were 20 right-handed normal volunteers (13 males and seven females, age range 24–64 years). For the latter task, control subjects were 10 right-handed normal volunteers (five females and five males, age range 22–56 years).
All subjects gave informed consent to participate in the study which was approved by The Ethics Committee of The Kantonspital, St Gallen.
Behavioural tasks
Tactile recognition
The blindfolded patients assessed 30 everyday objects by tactile exploration. As to size, the objects were deliberately chosen to allow convenient tactile exploration (i.e. a walnut, a peg, a key). Objects that were not correctly identified tactually were presented again to the subjects during three specific matching tasks. In each of these tasks, the replica for 10 objects had to be found in an array of five varying objects. (i) Tactile–tactile matching: the blindfolded patients had to recognize tactually an object previously not identified by tactile exploration. It should be noted that tactile–tactile matching was done exclusively by the affected hand, i.e. there was no contribution from the hemisphere contralateral to the lesion. (ii) Tactile–visual matching: the patients had to recognize visually an object previously not identified by tactile exploration. (iii) Visual imagination–tactile matching: the patients were asked to imagine a specific object previously not identified by tactile exploration, and to recognize this object by tactile exploration. The order of groups of objects as well as the order of objects within groups were pseudorandomized. The maximum time allowed for tactile exploration of an object, visual inspection of a group of objects, and visual imagination of an object, respectively, was 10 s.
Somatosensory discrimination of length differences and roughness
We employed a slight alteration of a paradigm for discrimination of oblongness, which originally has been described by Roland and Mortensen (1987). In keeping with the original paradigm, we used tactile stimuli consisting of pairs of parallelepipeds: 95% pure aluminium, identical volume (11.5 cm3), mass and surface qualities, but of different oblongness. We refined the method somewhat in that we used a Bernoulli model and logistic regression in order to define normative data and thresholds for conscious perception of length differences (Weder et al., 1998). In this study, two pairs of parallelepipeds were used for tactile exploration, i.e. cuboids with differences of 3.97 and 0.51 mm in the long axes, the former lying above and the latter below the perception threshold of 
2 mm as defined in normal volunteers; differences in the square bases were below threshold, i.e. 0.99 and 0.2 mm, in each pair (Weder et al., 1998). The blindfolded patients were asked to indicate the more oblong cuboid of a pair following sequential tactile exploration. Within a stimulation sequence, 10 pairs of cuboids were consecutively presented with pseudorandom differences either above or below the discrimination threshold. The order of presentation of the cuboids within a pair was also pseudorandomized. The exploration time was restricted to 5 s for each object, followed by a pause of 5 s. The answer, i.e. explicitly naming the object recognized, was given verbally by the subjects immediately after having explored the second cuboid of a pair. Several runs were performed with 40 explorations of each pair. A representative stimulation sequence was recorded for each patient on videotape, which afforded analysis of finger movements during tactile exploration.
The paradigm for roughness discrimination has been previously described in detail by Morley et al. (1983). In brief, the stimuli are surfaces with different grating profiles consisting of alternating grooves and ridges. The scale of roughness is defined by the spatial period length. Moving the pad of index finger back and forth, subjects were able to recognize differences of 10% from a reference spatial period length of 1000 µm at a probability of being correct (PC) of 0.9. In contrast, differences of 5% were correctly assessed at a PC of 0.75 (Morley et al., 1983). In the present study, subjects explored with their finger-pads stimuli consisting of synthetic surfaces with spatial period lengths of 1000 and 1100 µm. The contact pressure was kept constant by a counterbalancing weight of 150 g. Subjects were asked to indicate the rougher of the two surface gratings. Time restrictions, pseudorandomized order of grating profiles and verbal response followed the rules described for the length discrimination task (see above). Six runs, consisting of 10 comparisons, were performed. A representative stimulation sequence was recorded for each patient on videotape to analyse the exploration pattern, i.e. the frequency and average velocities of the sinusoidal finger movements. The contact area of the finger-pad was measured in order to estimate the average contact pressure.
Neuroanatomical assessment
The presence and precise localization of neuroanatomical deficits in the patients was assessed with a 1.5 Tesla MRI system (Symphony; Siemens, Erlangen). One millimetre thick contiguous slices were acquired in a sagittal plane using a T1-weighted gradient–echo sequence (Flash 3d; repetition time, TR = 1960 ms; echo time, TE = 3.93 ms; flip angle, 
= 15°; inversion time, TI = 1100 ms). Voxel size was 0.8 x 0.8 x 1 mm, allowing secondary multiplanar reconstructions. Acquisition time was 8 min 23 s.
Data analyses
Behavioural performance
Performance of tactile recognition and the matching tasks for each subject were expressed as the ratio between the number of correct answers and the total number of trials. Both intramodal matching and identification after visual information transfer were compared with tactile exploration alone by the odds ratio (OR).
Since somatosensory discrimination tasks exhibit a binomial character, i.e. the answer is either right (1) or wrong (0), we used a Bernoulli model and calculated the necessary parameters from the sample. Observations in normals and patients were approximated to a normal distribution from which a 95% confidence interval (CI) could be determined. In tasks with a binomial structure, the threshold for conscious perception is taken to be a PC of 0.75. We recently confirmed this threshold in the case of length perception (Weder et al., 1998). Thus, in the present study, a PC of >0.75 was presumed to reflect explicit information processing. In the length discrimination task, a pair with a length difference of 0.51 mm in the long axis served as reference for we previously demonstrated that this length difference yielded recognition responses after tactile exploration at chance level only (Weder et al., 1998). Task performances for the patients were compared by calculating the OR for both length and roughness discrimination. For both, we referred to our normative data as mentioned above. Patients performance was assessed with two-tailed, one sample analysis in which significance was assumed at P < 0.05.
Neuroanatomical assessment—MRI analysis
The extent of lesions in the patients was determined by high-resolution MRIs. The images were introduced into the computerized brain atlas (CBA) (Greitz et al., 1991; Thurfjell et al., 1995). This programme includes the interactive adaptation of the atlas standard brain to the patient’s anatomy by adjusting linear and non-linear parameters. Quality of the fitting procedure can be verified for any structure in the anatomic data bank of CBA. Coordinates of the CBA can be translated into the stereotaxic space of Talairach since spatial dimensions of the atlas standard brain and the spatial orientation of its intercommissural line have been determined (Gulyás and Roland, 1991). Concerning the hand area, the planes were chosen with reference to the activation fields as delineated by functional MRI (fMRI) of tactile exploration (Boecker et al., 1995). Furthermore, lesions were assigned rather to Brodmann area (BA) 3a and 3b or BA 1 and 2, depending on the position in the depth of central sulcus or the top of the postcentral gyrus.
|
Results |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Behavioural performance
Tactile recognition
Clinical data of the patients are summarized in Table 1. In Patient 1, tactile object recognition by the right hand was almost completely lost, as only one object was recognized out of 30. Impairment of object recognition by the right hand in Patient 2 was partial but still severe, as only nine objects were correctly identified out of 30. Moreover, this patient’s responses were substantially delayed (up to 10 s) compared with controls. In both patients, object recognition by the left hand was complete and immediate. Primary sensory processing—specifically, light touch, position sense and vibration sense of fingers—was normal in both individuals, but two-point discrimination was impaired (see Table 1). In Patient 1, the pattern of tactile exploration by the right hand was characterized by slow and irregular scanning movements with some enlarged trajectories. In Patient 2, grip force and finger-tapping were reduced or slowed, respectively, relative to the unaffected left hand. His exploratory movements were clumsy due to the interfering spasticity, and the pattern of tactile exploration was predominated by encompassing.
The findings of the intra- and transmodal matching tasks are summarized in Table 2. Shown are the results obtained exclusively with objects previously not identified by tactile exploration. In the tactile–tactile matching task performed sequentially with the affected hand only, the patients were able to identify the replicas of a considerable number of objects, although they did not recognize any object explicitly. The tactile–visual and the visual imagination–tactile matching tasks demonstrated a high proportion of recognition of tactually not recognized objects. This improvement in implicit object recognition (tactile–tactile matching) or in explicit recognition of objects (tactile–visual and visual imagination–tactile matching) was significant in comparison with tactile exploration alone.
|
Somatosensory discrimination of length differences and roughness
The task performance of the patients is summarized in Table 3.
|
During discrimination of length, exploring thumb movements (which are most important in this sort of task; see Seitz et al., 1991
) were slowed in the patients, in both the affected and the non-affected hands. In particular, exploring movements in Patient 2 were very slow on the right-hand side. Interestingly, despite this lower rate of exploration, this patient’s discrimination capacity was practically preserved, as indicated by a score falling within the lower range of normal volunteers. In contrast, although showing only moderately reduced finger frequency rates for exploration, Patient 1 failed to reach normal levels of object discrimination while using the affected right hand. Her score was lower than the critical threshold of 0.75 for conscious discrimination and was significantly reduced compared with normal volunteers. The trajectories of her finger movements were enlarged and less precisely adapted to the objects than might be expected in normal volunteers. Using the left hand, both patients explored at moderately reduced finger frequencies, but with discrimination abilities comparable to those of normal volunteers using the dominant right hand.
In the roughness discrimination task, frequency and average velocity of scanning movements exerted by the forefinger were generally slower for the patients than for normal volunteers. In Patient 1, the frequency of forefinger movements was 1.2 and 1.5 Hz on the right- and left-hand side, respectively, corresponding to an average velocity of 88 and 110 mm/s, respectively. The respective values in Patient 2 were 0.6 and 1 Hz, or 44 and 73.3 mm/s, respectively. Values for normal controls for frequency and average velocity on the right side were 1.8 ± 0.5 Hz and 129 ± 37 mm/s, respectively. For all individuals, the average contact area of the scanning fingerpad was 123 mm2 and the average contact pressure 0.6 g x wt x mm2, which fall within the range reported by Morley et al. (1983). The adaptation of the finger pad during the task was adequate in the patients, since the controlling weight could be counterbalanced fairly well. On the affected side, the discrimination of roughness was significantly reduced in both patients compared with normal volunteers. The proportion of correct answers for the patients was clearly below the critical threshold for conscious discrimination. On the left side, the performance was in the range of normal volunteers performing the task with the right hand.
The OR described the relationship of somatosensory discriminations between the two patients as follows: the macrogeometric property, the difference in length of 3.97 mm between two parallelepipeds, was better recognized by Patient 2 with a factor of 3.86 (95% CI 1.19–12.5). For comparison, the OR for the sub-threshold length difference of 0.51 mm was 0.86 (95% CI 0.4–1.81). The microgeometric property, the difference in length period of a grating profile as described above, was poorly recognized by both patients, as reflected by an OR of 0.9 (95% CI 0.41–2.01) (Fig. 1).
|
l) of 3.97 mm. In contrast, both patients failed to the same degree in recognition of microgeometric properties, i.e. in the discrimination of length differences of periods in a grating profile (difference of a spatial period of 1000 µm).Neuroanatomical assessment—MRI data
Figure 2 summarizes the main lesions as depicted by high resolution cranial MRIs. Note that in Patient 1, the lesion involved mainly the posterior bank of postcentral gyrus (i.e. BA 1 and 2; Figs 2A and B, and 3A). In addition, the lesion extended dorsomedially to the left superior parietal lobule and dorsolaterally to the supramarginal gyrus; the anterior part and the cortex lining the intraparietal sulcus were especially included within the lesion (Figs 2A and B, and 3A). Ventrolaterally, the lesion affected the parietal operculum, which corresponds to the secondary somatosensory cortex (SII), and the left retroinsular cortex (Fig. 2C). The angular gyrus was intact. In contrast, in Patient 2 the lesion was rather limited and involved the postcentral gyrus and part of the primary motor cortex. The anterior bank of the postcentral gyrus was damaged (i.e. BA 3a and 3b) at the sulcal depth (Figs 2D and E, and 3B). On the lateral border, this area touches SII. The superior parietal lobule as well as the inferior parietal regions, such as the supramarginal gyrus, were not affected. The vascular lesion extended deeply into the white matter, barely bypassing the parietal operculum (Fig. 2F).
|
|
|
Discussion |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
The two cases showed severe unilateral aperceptive tactile agnosia. However, the perception of underlying micro- and macrogeometrical properties by the affected hand was different in each case. Specifically, discrimination of roughness was impaired in both patients, but length discrimination was disturbed in only one case. These results demonstrate that tactile recognition is differentially affected by different classes of object properties. The patients found the replica of more objects in a tactile–tactile matching task than expected from tactile exploration alone, revealing an enhanced ability for implicit object recognition. Moreover, explicit recognition of objects was enhanced significantly in both patients when complementary extra-modal information was available. The morphological lesions in both patients involved the left postcentral gyrus and extended in the more severe case to the parietal operculum (i.e. SII), and the supramarginal gyrus and posterior superior parietal lobule. Taken together, these findings point to multiple critical sites in the parietal cortex that interfered with perception during normal tactile object exploration (Fig. 4).
|
Terminological considerations
In the literature, there is little consistency in the terminology used to designate a disturbance in tactile object recognition. Astereognosis, the term introduced by Hoffmann (1885
), was understood to denote a loss of tactile object recognition due to impaired tactile spatial perception. Disturbed perception of macrogeometrical features such as the size and shape of objects was supposed by Roland (1976) to be central to a disturbance in tactile object recognition. However, referring to an early report by Delay (1935), Mauguiere et al. (1983) proposed that astereognosis might represent a more complex perceptual disorder, i.e. defective perception of both microgeometrical and macrogeometrical object properties. Thus, until recently the designations astereognosis and tactile agnosia were used interchangeably, with no definition of at which levels an abnormality in information processing occurs: elementary sensory (sensation), intermediate (perception) and/or higher (recognition) (Caselli, 1997). How ever, tactile agnosia occurred in comparable case studies in pure form, i.e. the associated elementary sensory disorder was too minor to be the source of the failure (Mauguiere et al., 1983; Endo et al., 1992; Caselli, 1993; Reed and Caselli, 1994; Mauguiere and Isnard, 1995; Platz, 1996; Reed et al., 1996). This suggests that the concept should be reserved exclusively for disturbed intermediate and higher order sensory processing. As demonstrated previously (Klatzky et al., 1987), subjects group explored objects tactually mainly by texture and intrinsic object properties like hardness, and only to a lesser extent by shape. This is in clear contrast to visual classification and implies that astereognosis should be restricted to impaired tactile spatial perception (Caselli, 1997). Correspondingly, degraded sensory processing of microgeometrical features was a powerful determinant in our patients (see below).
Role of micro- and macrogeometrical features for object recognition
Both patients in this study were significantly deficient in tactile object recognition. However, the precise character of this deficiency was remarkably different; although neither patient perceived differences in surface roughness beyond threshold, one patient was able to extract length differences among three dimensional objects and, thus, explicitly differentiate parallelepipeds with respect to this criterion.
Tactile object recognition occurs almost immediately in normal volunteers; both patients exhibited comparable performance when using the unaffected hand. Several objects or intrinsic object properties were recognized by Patient 2 after considerable latency, and he obviously tried to complete real associations by integrating the partial aspects recognized. We hypothesize that this disintegration from whole-object recognition to recognition of partial aspects in Patient 2, and to complete failure to recognize the object at all in Patient 1, reflects differentially degraded and, thus, distinct dysfunctional networks in the parietal cortex for recognition of the micro- and macrogeometrical properties of objects. Thus, we suggest that the hypothesized networks are necessary for normal tactile object recognition.
Roughness discrimination most probably facilitates recognition of textures, which is of importance for the classification of objects (see above; Klatzky et al., 1987; Caselli, 1997), e.g. the recognition of embossed letters studied by Phillips et al. (1988). This explains the severely disturbed tactile object recognition in both patients. A further, possibly less important aspect of feature extraction during tactile exploration is the spatial assessment of objects on a macrogeometrical level (Binkofski et al., 2001). It seems that the tactile agnosia in Patient 1 was characterized by this additional perceptual deficit.
Improved explicit object recognition after transmodal information transfer
Tactile agnosia may be subdivided into an aperceptive and an associative form (Mauguiere and Isnard, 1995; Platz, 1996; Reed et al., 1996; Mesulam, 1998). Consequently, aperceptive tactile agnosia signifies a defective modality-specific perception due to inappropriate extraction of features from the objects explored. In this regard, perception may be intact, but there may be no access to the resulting percept (Tranel, 1991; Platz, 1996). The corresponding failure is assumed to represent a modality-specific associative tactile agnosia. There has been no experimental proof for such a differentiation, since the exact transition from perception to recognition is difficult to assess (Damasio et al., 1992). Based on indirect evidence, however, rare cases of disturbed tactile object recognition were classified both as aperceptive tactile agnosias (Mauguiere et al., 1983; Caselli, 1993; Reed and Caselli, 1994; Mauguiere and Isnard, 1995; Reed et al., 1996) and associative tactile agnosias (Endo et al., 1992; Platz, 1996; Nakamura et al., 1998). Such classifications proceeded from an assessment of clinical and experimental test profiles that indicated either intact or defective tactile object perception. However, associative tactile agnosia is thought by other researchers to arise only when transmodal access to established knowledge of objects is disrupted (Mesulam, 1998).
A long history of controversy surrounds this separation into modality-specific and supra- or extra-modal associations during the recognition process (Damasio et al., 1992). Our two patients showed evidence of a primary impairment of perception by virtue of reduced recognition of defined object properties as shown by the somatosensory discrimination tasks for length differences and roughness. We suggest that this impairment reflected the lack of a valid percept of real-world objects for access to semantic memory. A transmodal information transfer from or into the defective specific system for perception (here, from visual memory to tactile exploration) may afford successful explicit recognition of objects. We hypothesize that this complementary information, which is based primarily on an intact visual percept or knowledge about it, became operative by evoking and synchronizing neural activity in local and non-local convergence zones (Damasio, 1989a). Indeed, enhanced neuronal activity has been shown within visual cortical areas of blind humans after cross-modal plasticity (Cohen et al., 1997). In this regard, we conclude that the residual capacity for correct allocation of objects explored, i.e. in the tactile–tactile matching task, suggests implicit processing of specific sensory information within the perceptual representation system (Tulving and Schacter, 1990). Moreover, we suggest that this implicit level of perceptual processing was the necessary foundation for further explicit recognition processing. Thus, we propose that explicit object recognition occurs when an implicitly formed percept passes the threshold for recognition of an object after a cross-modal comparison.
MRI lesion analysis
MRI lesion analysis revealed three areas of neuroanatomical damage in the two patients studied here: (i) the postcentral gyrus (SI); (ii) SII or retroinsular/parietal operculum complex; and (iii) the posterior parietal lobe, including the supramarginal gyrus and superior parietal lobule. We hypothesize that each of these areas represents a relay node of a distributed network subserving normal tactile object recognition.
Postcentral gyrus (SI)
Both patients recognized simple sensory stimuli such as light touch, pain, temperature, position sense of fingers and vibration, suggesting normal early sensory processing. However, both patients found difficulty with increasing complexity of response properties, i.e. two-point discrimination and discrimination of grating profiles by moving the exploring finger-pad back and forth. In terms of receptive sub-fields, neurones involved in the input of simple punctate stimuli respond preferentially in BA 3a and 3b of the postcentral gyrus, whereas neurones in subsequent cortical processing stations, BA 1 and BA 2, exhibit complex response properties (Hyvärinen and Poranen, 1978; Iwamura et al., 1985; Gardner et al., 1989). Consequently, feature detection (such as texture properties) does not arise from thalamic input, but from cortical processing of more elementary inputs (Phillips et al., 1988). Hence, certain neurones in BA 2 receive input from more than one sub-modality, e.g. touch and position. In accordance with this concept, BA 1 and 2 are involved mainly in the deficits observed in Patient 1. Patient 2 presented a different lesion pattern; we suggest that BA 1 and 2 were reduced to processing simple sensory stimuli due to deafferentiation from lesioned BA 3a and 3b. Similarly, SI damage has been shown to impair discriminative sensations, particularly two-point discrimination and stereognosis, leaving superficial sensation spared (Corkin et al., 1970; Roland, 1976).
Secondary sensory area (SII)
The extension of the lesion to parietal operculum, involving SII and including the retroinsular cortex in Patient 1 probably affected tactile object recognition further, since the parietal operculum/retroinsular complex seems to be specifically involved in the perception of intrinsic object properties and associated microgeometrical features (O’Sullivan et al., 1994; Ledberg et al., 1995; Roland et al., 1998; Binkofski et al., 1999a). In Patient 2, the lesion extended deeply into the neighbouring white matter, just bypassing the parietal operculum. Thus, as in Patient 1, the lesion in Patient 2 probably disrupted significantly cortico-cortical interconnections between SI and parietal operculum. In contrast to Patient 1, however, this patient’s ability to recognize intrinsic object properties was partially preserved. Recent neurophysiological data in humans have suggested an important role for SII in later stages of cortical processing, indicating that this region responds to contralateral sensory stimuli via SI (Forss et al., 1999). Thus, it has been proposed that SI and SII are serially organized in humans and, at least in primates, there is tonic facilitation of SII by SI (Zhang et al., 1996). It cannot be clearly ascertained whether the disruption of this ventrolateral somatosensory association cortex (Caselli, 1993) affected the primarily degraded percept in our patients beyond tactile association.
Posterior parietal lobe
Lesion and functional imaging studies have verified the significance of superior parietal lobule for the perception of macrogeometrical object features (O’Sullivan et al., 1994; Ledberg et al., 1995; Binkofski et al., 1998; Roland et al., 1998). In animal experiments, the neurones of this area have been shown to be particularly responsive to passive joint and active limb movements (Sakata et al., 1973; Mountcastle et al., 1975; Lacquaniti et al., 1995). Both the superior parietal lobule and the supramarginal gyrus may be critically involved in somatosensory discrimination. Indeed, PET studies have revealed that these regions are activated during length discrimination tasks (O’Sullivan et al., 1994) and, correspondingly, deactivated when length discrimination is impaired (Weder et al., 2000). Bodegård et al. (2001) established a critical role of mainly the anterior part of supramarginal gyrus and the cortex lining intraparietal sulcus in active and passive shape discrimination. In addition, the supramarginal gyrus plays an important role in the coordination of fine finger movements required for object manipulation, as shown recently in an fMRI task (Binkofski et al., 1998). Owing to the dense interconnections with the lateral premotor cortex (Strick and Kim, 1978; Petrides and Pandya, 1984), lesions of the superior and inferior parietal cortex disturb sensory guided movements and result in tactile apraxia. This parieto-premotor network for object manipulation has been validated in fMRI studies (Binkofski et al., 1999a, b). We suggest that an interruption in such a network explains the degradation of exploratory movements in Patient 1. Motor performance in this case corresponded to tactile apraxia, a unimodal disturbance of exploratory motor function (Binkofski et al., 2001). In contrast, Patient 2 had an additional primary motor dysfunction that did not interfere with the extraction of macrogeometrical information from the objects. In short, unlike Patient 2, Patient 1 was deprived of macrogeometrical information. This more severe deficit was reflected in the additional lesion effects of the anterior and posterior parietal cortex observed in Patient 1, because BA 5 and 7 are organized parallel to SI, with direct projections from the thalamus, which is in contrast to the organizational structure of SII (Forss et al., 1999). The main output of the lateral posterior parietal cortices seems to be directed to supplementary sensory area, i.e. the dorsomedial association cortex (Cavada et al., 1989), which was intact in our patients.
Our findings concur well with those of previous lesion studies and support our hypothesis of distributed networks subserving sensory processing of tactile recognition, i.e. lesions of the postcentral gyrus and associated SII are now thought to be involved in most cases of tactile agnosia (Corkin et al., 1970; Roland, 1976). Moreover, more recent studies of the disturbance have suggested a critical role for the superior (Pause et al., 1989; Freund, 1995; Binkofski et al., 2001) or the inferior parietal lobe (Caselli, 1991).
Conclusion
In summary, our results demonstrate perceptual disturbances underlying tactile agnosia that involve deficits at both implicit and explicit levels of information processing. Importantly, these findings suggest that normal tactile information processing is subserved by distributed networks in the parietal cortex, and that hierarchical and parallel processing of complementary information relies on the integrity of the postcentral gyrus and posterior parietal lobe. Specifically, we found further evidence for dual streams of somatosensory information processing flowing to the ventrolateral and dorsomedial somatosensory association cortex, respectively (Caselli, 1993). Future work may provide further insight into the behavioural, neuroanatomical and neurophysiological correlates of normal tactile perceptual processes.
|
Acknowledgements |
|---|
We wish to thank Drs Nina Azari and John Missimer for editing the manuscript, and Professor Alex Keel of the University of St Gallen for help in statistical analysis of the data.
|
References |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Bassetti C, Bogousslavsky J, Regli F. Sensory syndromes in parietal stroke. Neurology 1993; 43: 1942–9.
Binkofski F, Dohle C, Posse S, Stephan KM, Hefter H, Seitz RJ, et al. Human anterior intraparietal area subserves prehension. Neurology 1998; 50: 1253–9.
Binkofski F, Buccino G, Posse S, Seitz RJ, Rizzolatti G, Freund H. A fronto-parietal circuit for object manipulation in man: evidence from an fMRI-study. Eur J Neurosci 1999a; 11: 3276–86.[Web of Science][Medline]
Binkofski F, Buccino G, Stephan KM, Rizzolatti G, Seitz RJ, Freund HJ. A parieto-premotor network for object manipulation: evidence from neuroimaging. Exp Brain Res 1999b; 128: 210–3.[Web of Science][Medline]
Binkofski F, Kunesch E, Classen J, Seitz RJ, Freund HJ. Tactile apraxia. Unimodal apractic disorder of tactile object exploration associated with parietal lobe lesions. Brain 2001; 124: 132–44.
Bodegård A, Geyer S, Grefkes C, Zilles K, Roland PE. Hierarchical processing of tactile shape in the human brain. Neuron 2001; 31: 317–28.[Web of Science][Medline]
Boecker H, Khorram-Sefat D, Kleinschmidt A, Merboldt K, Hänicke W, Requardt M, et al. High-resolution functional magnetic resonance imaging of cortical activation during tactile exploration. Hum Brain Mapp 1995; 3: 236–44.[Web of Science]
Caselli RJ. Rediscovering tactile agnosia. [Review.] Mayo Clin Proc 1991; 66: 129–42.[Web of Science][Medline]
Caselli RJ. Ventrolateral and dorsomedial somatosensory association cortex damage produces distinct somesthetic syndromes in humans. Neurology 1993; 43: 762–71.
Caselli RJ. Tactile agnosia and disorders of tactile perception. In: Feinberg TE, Farah MJ, editors. Behavioral neurology and neuropsychology. New York (NY): McGraw-Hill; 1997. p. 277–88.
Cavada C, Goldman-Rakic PS. Posterior parietal cortex in rhesus monkey: I. Parcellation of areas based on distinctive limbic and sensory corticocortical connections. J Comp Neurol 1989; 287: 393–421.[Web of Science][Medline]
Cohen LG, Celnik P, Pascual-Leone A, Corwell B, Falz L, Dambrosia J, et al. Functional relevance of cross-modal plasticity in blind humans. Nature 1997; 389: 180–3.[Medline]
Corkin S, Milner B, Rasmussen T. Somatosensory thresholds: contrasting effects of postcentral-gyrus and posterior parietal-lobe excisions. Arch Neurol 1970; 23: 41–58.
Damasio AR. The brain binds entities and events by multi regional activation from convergence zones. Neural Comput 1989; 1: 123–32.
Damasio AR, Tranel D, Eslinger P. The agnosias. In: Asbury AK, McKhann GM, McDonald WI, editors. Diseases of the nervous system: clinical neurobiology, Vol. 1. 2nd ed. Philadelphia (PA): W.B. Saunders; 1992. p. 741–50.
Delay JP. Les astereognosies: pathologie du toucher, clinique, physiologie, topographie. Paris: Masson; 1935.
Endo K, Miyasaka M, Makishita M, Yanagisawa N, Sugishita M. Tactile agnosia and tactile aphasia: symptomatological and anatomical differences. Cortex 1992; 28: 445–69.[Web of Science][Medline]
Forss N, Hietanen M, Salonen O, Hari R. Modified activation of somatosensory cortical network in patients with right-hemisphere stroke. Brain 1999; 122: 1889–99.
Freund HJ. The apraxias. In: Kennard C, editor. Recent advances in clinical neurology, No. 8. Edinburgh: Churchill Livingstone; 1995. p. 29–49.
Gardner EP, Hämäläinen HA, Palmer CI, Warren S. Touching the outside world: representation of motion and direction within primary somatosensory cortex. In: Lund JS, editor. Sensory processing in the mammalian brain: neural substrates and experimental strategies. New York (NY): Oxford University Press; 1989. p. 49–66.
Greitz T, Bohm C, Holte S, Eriksson L. A computerized brain atlas: construction, anatomical content, and some applications. J Comput Assist Tomogr 1991; 15: 26–38.[Web of Science][Medline]
Gulyás B, Roland PE. Cortical fields participating in form and colour discrimination in the human brain. Neuroreport 1991; 2: 585–8.[Web of Science][Medline]
Hoffmann H. Stereognostische Versuche, angestellt zur Ermittlung der Elemente des Gefühlsinnes, aus denen die Vorstellungen des Körpers im Raume gebildet werden. Dt Arch Klin Med 1885; 36: 398–426.
Hyvärinen J, Poranen A. Movement-sensitive and direction and orientation-selective cutaneous receptive fields in the hand area of the post-central gyrus in monkeys. J Physiol 1978; 283: 523–37.
Iwamura Y, Tanaka M, Sakamoto M, Hikosaka O. Comparison of the hand and finger representation areas 3, 1, and 2 of the monkey somatosensory cortex. In: Rowe M, Willis WD Jr, editors. Development, organization, and processing in somatosensory pathways. New York (NY): Liss; 1985. p. 253–67.
Klatzky RL, Lederman S, Reed C. There’s more to touch than meets the eye: the salience of object attributes for haptics with and without vision. J Exp Psychol Gen 1987; 116: 356–69.[Web of Science]
Lacquaniti F, Guigon E, Bianchi L, Ferraina S, Caminiti R. Representing spatial information for limb movement: role of area 5 in the monkey. Cereb Cortex 1995; 5: 391–409.
Ledberg A, O’Sullivan BT, Kinomura S, Roland PE. Somatosensory activations of the parietal operculum of man. Eur J Neurosci 1995; 7: 1934–41.[Web of Science][Medline]
Mai JK, Assheuer J, Paxinos G. Atlas of the human brain. San Diego (CA): Academic Press; 1997.
Mauguiere F, Isnard J. Tactile agnosia and dysfunction of the primary somatosensory area. Data of the study by somatosensory evoked potentials in patients with deficits of tactile object recognition [French]. Rev Neurol (Paris) 1995; 151: 518–27.[Medline]
Mauguiere F, Desmedt JE, Courjon J. Astereognosis and dissociated loss of frontal or parietal components of somatosensory evoked potentials in hemispheric lesions. Brain 1983; 106: 271–311.
Mesulam M. From sensation to cognition. [Review.] Brain 1998; 121: 1013–52.
Morley JW, Goodwin AW, Darian-Smith I. Tactile discrimination of gratings. Exp Brain Res 1983; 49: 291–9.[Web of Science][Medline]
Mountcastle VB, Lynch JC, Georgopoulos A, Sakata H, Acuna C. Posterior parietal association cortex of the monkey: command functions for operations within extrapersonal space. J Neurophysiol 1975; 38: 871–908.
Nakamura J, Endo K, Sumida T, Hasegawa T. Bilateral tactile agnosia: a case report. Cortex 1998; 34: 375–88.[Web of Science][Medline]
O’Sullivan BT, Roland PE, Kawashima R. A PET study of somatosensory discrimination in man. Microgeometry versus macrogeometry. Eur J Neurosci 1994; 6: 137–48.[Web of Science][Medline]
Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 1971; 9: 97–113.[Web of Science][Medline]
Pause M, Kunesch E, Binkofski F, Freund HJ. Sensorimotor disturbances in patients with lesions of the parietal cortex. Brain 1989; 112: 1599–625.
Petrides M, Pandya DN. Projections to the frontal cortex from the posterior parietal region in the rhesus monkey. J Comp Neurol 1984; 288: 105–16.
Phillips JR, Johnson KO, Hsiao SS. Spatial pattern representation and transformation in monkey somatosensory cortex. Proc Natl Acad Sci USA 1988; 85: 1317–21.
Platz T. Tactile agnosia. Casuistic evidence and theoretical remarks on modality-specific meaning representations and sensorimotor integration. Brain 1996; 119: 1565–74.
Qureshi AI, Luft AR, Sharma M, Guterman LR, Hopkins LN. Prevention and treatment of thromboembolic and ischemic complications associated with endovascular procedures. [Review.] Neurosurgery 2000; 46: 1360–75.[Web of Science][Medline]
Reed CL, Caselli RJ. The nature of tactile agnosia: a case study. Neuropsychologia 1994; 32: 527–39.[Web of Science][Medline]
Reed CL, Caselli RJ, Farah MJ. Tactile agnosia. Underlying impairment and implications for normal tactile object recognition. Brain 1996; 119: 875–88.
Roland PE. Astereognosis. Tactile discrimination after localized hemispheric lesions in man. Arch Neurol 1976; 33: 543–50.
Roland PE. Somatosensory detection of microgeometry, macrogeometry and kinaesthesia after localized lesions of the cerebral hemispheres in man. [Review.] Brain Res 1987; 434: 43–94.[Medline]
Roland PE, Mortensen E. Somatosensory detection of microgeometry, macrogeometry and kinesthesia in man. [Review]. Brain Res 1987; 434: 1–42.[Medline]
Roland PE, O’Sullivan B, Kawashima R. Shape and roughness activate different somatosensory areas in the human brain. Proc Natl Acad Sci USA 1998; 95: 3295–300.
Sakata H, Takaoka Y, Kawarasaki A, Shibutani H. Somatosensory properties of neurons in the superior parietal cortex (area 5) of the rhesus monkey. Brain Res 1973; 64: 85–102.[Web of Science][Medline]
Seitz RJ, Roland PE, Bohm C, Greitz T, Stone-Elander S. Somatosensory discrimination of shape: tactile exploration and cerebral activation. Eur J Neurosci 1991; 3: 481–92.[Web of Science][Medline]
Strick PL, Kim CC. Input to primate motor cortex from posterior parietal cortex (area 5). I. Demonstration by retrograde transport. Brain Res 1978; 157: 325–30.[Web of Science][Medline]
Talairach J, Tournoux P. Co-planar stereotaxic atlas of the human brain. 3-dimensional proportional system. An approach to cerebral imaging. Stuttgart: Thieme; 1988.
Thurfjell L, Bohm C, Bengtsson E. CBA – an atlas-based software tool used to facilitate the interpretation of neuroimaging data. Comput Methods Programs Biomed 1995; 47: 51–71.
Tranel D. What has been rediscovered in ‘rediscovering tactile agnosia’? Mayo Clin Proc 1991; 66: 210–4.[Web of Science][Medline]
Tulving E, Schacter DL. Priming and human memory systems. Science 1990; 247: 301–6.
Weder B, Nienhusmeier M, Keel A, Leenders KL, Ludin HP. Somatosensory discrimination of shape: prediction of success in normal volunteers and parkinsonian patients. Exp Brain Res 1998; 120: 104–8.[Web of Science][Medline]
Weder B, Azari NP, Knorr U, Seitz RJ, Keel A, Nienhusmeier M, et al. Disturbed functional brain interactions underlying deficient tactile object discrimination in Parkinson’s disease. Hum Brain Mapp 2000; 11: 131–45.[Web of Science][Medline]
Wiebers DO, Dale AJD, Kokmen E, Swanson JW. The sensory examination. In: Wiebers DO, Dale AJD, Kokmen E, Swanson J, editors. Mayo Clinic examinations in neurology. 7th ed. St Louis (MO): Mosby; 1998. p. 255–74.
Zhang HQ, Murray GM, Turman AB, Mackie PD, Coleman GT, Rowe MJ. Parallel processing in cerebral cortex of the marmoset monkey: effect of reversible SI inactivation on tactile responses in SII. J Neurophysiol 1996; 76: 3633–55.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  L. B. Hinkley, L. A. Krubitzer, S. S. Nagarajan, and E. A. Disbrow Sensorimotor Integration in S2, PV, and Parietal Rostroventral Areas of the Human Sylvian Fissure J Neurophysiol, February 1, 2007; 97(2): 1288 - 1297. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-O. Karnath and M.-T. Perenin Cortical Control of Visually Guided Reaching: Evidence from Patients with Optic Ataxia Cereb Cortex, October 1, 2005; 15(10): 1561 - 1569. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Schwartz, F. Assal, N. Valenza, M. L. Seghier, and P. Vuilleumier Illusory persistence of touch after right parietal damage: neural correlates of tactile awareness Brain, February 1, 2005; 128(2): 277 - 290. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||