wiki:XbrainEvaluation

Version 37 (modified by detwiler, 12 years ago) (diff)

--

Answering Some Real-World Questions


Questions:

David Corina recommended the following "simple" queries for DXBrain:

1) Do males or females have more of a tendency to give semantic paraphasias?
2) Are Male or females likely to have a greater spread of naming sites?
3) If a subject has a naming error in the territory of Broca's area, what are the next most likely associated regions?
4) Is there an age difference in the types of errors that are produced (young people provide "no-response" while old people produce "aphasia" etc.)?
5) Within a given subject, are there regions that give rise to phonological errors that do not also show a semantic error?
6) Within a given subject, are there regions that give rise to phonological reductions that do not also show a semantic error?
7) For the 3 cases that exhibited morphological errors, what were the intended targets and exact responses?
8) Within a given subject, are there regions that give rise to semantic errors that do not also show a phonological error?
9) Within a given subject, are there regions that give rise to semantic errors that do not also show a phonological reduction?
10) What error producing trials lie within a specified deviation of a target MNI coordinate?

Solutions:

1)
Results:

<results>
   <male_results>
      <surgery_patient_count>42</surgery_patient_count>
      <semantic_paraphasia_count>12</semantic_paraphasia_count>
      <percentage>28.57142857142857</percentage>
   </male_results>
   <female_results>
      <surgery_patient_count>46</surgery_patient_count>
      <semantic_paraphasia_count>19</semantic_paraphasia_count>
      <percentage>41.30434782608695</percentage>
   </female_results>
</results>

The previous results were generated by the following query: FirstQuery

2)
Results:

<results>
   <male>
      <avg_standard_deviation>74.22857388425521</avg_standard_deviation>
   </male>
   <female>
      <avg_standard_deviation>42.559016047797</avg_standard_deviation>
   </female>
</results>

The previous results were generated by the following query: SecondQuery

Note: This query also uses a functions that we implemented in a math xquery library: MathLibrary

The stategy employed for determining the "spread" of a patients naming sites was to calculate, for each patient, the standard deviation in the geometric coordinates of their associated naming sites and then calculating averages among the male and female patients. Note that the current values only consider type 2 and 6 error codes (semantic paraphasia and semantic/phonological blends) as naming errors. This query/result needs to be double checked and should likely include additional error types.

Note: This query considers only patients with more than one naming site. Patients with a single naming site were discarded. Also note that all patients' site coordinates were normalized before calculating the standard deviation.

If we include all error types from 2 to 10, the results are:

<results>
   <male>
      <avg_standard_deviation>48.46459431620677</avg_standard_deviation>
   </male>
   <female>
      <avg_standard_deviation>47.61985622347745</avg_standard_deviation>
   </female>
</results>

3) coming soon ...

4)
Results:

<results>
   <younger_patients>
      <error freq="51.52%">1</error>
      <error freq="20.76%">11</error>
      <error freq="6.42%">2</error>
      <error freq="6.34%">9</error>
      <error freq="4.60%">10</error>
      <error freq="1.99%">20</error>
      <error freq="1.73%">3</error>
      <error freq="1.65%">8</error>
      <error freq="1.39%">22</error>
      <error freq="1.21%">23</error>
      <error freq="1.12%">4</error>
      <error freq="0.52%">5</error>
      <error freq="0.26%">6</error>
      <error freq="0.26%">7</error>
      <error freq="0.17%">21</error>
   </younger_patients>
   <older_patients>
      <error freq="52.77%">1</error>
      <error freq="18.51%">11</error>
      <error freq="14.81%">10</error>
      <error freq="6.48%">2</error>
      <error freq="4.62%">5</error>
      <error freq="0.92%">9</error>
      <error freq="0.92%">6</error>
      <error freq="0.46%">3</error>
      <error freq="0.46%">4</error>
   </older_patients>
</results>

The previous results were generated by the following query: FourthQuery

For this query the threshold for dividing younger patients from older patients was chosen arbitrarily to be 50. However, other thresholds can be easily tested simply by changing the value of the $age_threshold variable at the beginning of the query. There appears to be some difference in the occurences of type 10 (off target) errors, although further analysis would be required to determine if this is statistically significant.

5)
Results:

<patient>
   <pnum>52</pnum>
   <region>
      <name>Middle part of middle temporal gyrus</name>
      <codes>
         <code>C</code>
         <code>1</code>
         <code>11</code>
         <code>FSE</code>
         <code>9</code>
         <code>!</code>
         <code>10</code>
         <code>B</code>
         <code>A</code>
         <code>3</code>
         <code>D</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>55</pnum>
   <region>
      <name>Anterior part of superior temporal gyrus</name>
      <codes>
         <code>!</code>
         <code>B</code>
         <code>3</code>
         <code>S</code>
      </codes>
   </region>
   <region>
      <name>Posterior part of supramarginal gyrus</name>
      <codes>
         <code>A</code>
         <code>3</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>57</pnum>
   <region>
      <name>Posterior part of middle frontal gyrus</name>
      <codes>
         <code>D</code>
         <code>5</code>
         <code>S</code>
         <code>FSE</code>
         <code>3</code>
         <code>!</code>
         <code>B</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>84</pnum>
   <region>
      <name>Middle part of middle temporal gyrus</name>
      <codes>
         <code>C</code>
         <code>1</code>
         <code>FSE</code>
         <code>B</code>
         <code>9</code>
         <code>S</code>
         <code>3</code>
         <code>!</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>91</pnum>
   <region>
      <name>Anterior part of middle temporal gyrus</name>
      <codes>
         <code>C</code>
         <code>B</code>
         <code>1</code>
         <code>3</code>
         <code>S</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>129</pnum>
   <region>
      <name>Anterior part of superior temporal gyrus</name>
      <codes>
         <code>3</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>139</pnum>
   <region>
      <name>Middle part of superior temporal gyrus</name>
      <codes>
         <code>B</code>
         <code>11</code>
         <code>23</code>
         <code>20</code>
         <code>!</code>
         <code>C</code>
         <code>S</code>
         <code>22</code>
         <code>3</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>141</pnum>
   <region>
      <name>Posterior part of superior temporal gyrus</name>
      <codes>
         <code>B</code>
         <code>1</code>
         <code>!</code>
         <code>3</code>
         <code>?</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>144</pnum>
   <region>
      <name>Middle part of superior temporal gyrus</name>
      <codes>
         <code>!</code>
         <code>3</code>
         <code>9</code>
         <code>S</code>
         <code>B</code>
         <code>1</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>152</pnum>
   <region>
      <name>Middle part of superior temporal gyrus</name>
      <codes>
         <code>C</code>
         <code>9</code>
         <code>S</code>
         <code>1</code>
         <code>23</code>
         <code>21</code>
         <code>B</code>
         <code>!</code>
         <code>20</code>
         <code>3</code>
         <code>11</code>
      </codes>
   </region>
   <region>
      <name>Middle part of middle temporal gyrus</name>
      <codes>
         <code>B</code>
         <code>C</code>
         <code>S</code>
         <code>!</code>
         <code>9</code>
         <code>23</code>
         <code>3</code>
         <code>11</code>
      </codes>
   </region>
</patient>

(Note: the above results have subsequently been augmented (query altered) to include tags consistent with the 2D result view and to include stimsite coordinates for the 3D result view)

The previous results were generated by the following query: FifthQuery

This question was interpreted as "give me all patients who exhibited a type 3 error (phonological) in a named anatomical region but did not exhibit a type 2 or 6 (semantic or phonological/semantic blend) in the same region, for each display the names and codes for all such regions".

6)
Results:

<patient>
   <pnum>55</pnum>
   <region>
      <name>Middle part of superior temporal gyrus</name>
      <codes>
         <code>B</code>
         <code>8</code>
         <code>11</code>
         <code>A</code>
         <code>1</code>
         <code>5</code>
         <code>FSE</code>
         <code>D</code>
         <code>9</code>
         <code>S</code>
         <code>!</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>60</pnum>
   <region>
      <name>Middle part of inferior temporal gyrus</name>
      <codes>
         <code>A</code>
         <code>8</code>
         <code>FSE</code>
         <code>R</code>
         <code>!</code>
         <code>C</code>
         <code>1</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>61</pnum>
   <region>
      <name>Middle part of middle temporal gyrus</name>
      <codes>
         <code>!</code>
         <code>8</code>
         <code>D</code>
         <code>10</code>
         <code>FSE</code>
         <code>?</code>
         <code>11</code>
         <code>1</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>62</pnum>
   <region>
      <name>Polar part of superior temporal gyrus</name>
      <codes>
         <code>B</code>
         <code>1</code>
         <code>FSE</code>
         <code>!</code>
         <code>A</code>
         <code>8</code>
         <code>S</code>
      </codes>
   </region>
   <region>
      <name>Opercular part of inferior frontal gyrus</name>
      <codes>
         <code>B</code>
         <code>11</code>
         <code>D</code>
         <code>1</code>
         <code>A</code>
         <code>8</code>
         <code>S</code>
      </codes>
   </region>
   <region>
      <name>Triangular part of inferior frontal gyrus</name>
      <codes>
         <code>B</code>
         <code>1</code>
         <code>11</code>
         <code>C</code>
         <code>8</code>
         <code>S</code>
         <code>FSE</code>
         <code>!</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>63</pnum>
   <region>
      <name>Anterior part of supramarginal gyrus</name>
      <codes>
         <code>8</code>
         <code>FSE</code>
         <code>!</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>84</pnum>
   <region>
      <name>Middle part of superior temporal gyrus</name>
      <codes>
         <code>C</code>
         <code>1</code>
         <code>B</code>
         <code>FSE</code>
         <code>11</code>
         <code>!</code>
         <code>8</code>
      </codes>
   </region>
</patient>
<patient>
   <pnum>87</pnum>
   <region>
      <name>Posterior part of supramarginal gyrus</name>
      <codes>
         <code>C</code>
         <code>S</code>
         <code>B</code>
         <code>11</code>
         <code>23</code>
         <code>21</code>
         <code>20</code>
         <code>8</code>
         <code>22</code>
         <code>10</code>
         <code>?</code>
      </codes>
   </region>
</patient>

(Note: the above results have subsequently been augmented (query altered) to include tags consistent with the 2D result view and to include stimsite coordinates for the 3D result view)

The previous results were generated by the following query: SixthQuery

This question was interpreted as "give me all patients who exhibited a type 8 error (phonological reduction) in a named anatomical region but did not exhibit a type 2 or 6 (semantic or phonological/semantic blend) in the same region, for each display the names and codes for all such regions".

7)
Results:

<patients>
   <patient>
      <pnum>63</pnum>
      <trial>
         <trial_num>95</trial_num>
         <item>mouse</item>
         <patient_response>"mice--mouse"</patient_response>
      </trial>
   </patient>
   <patient>
      <pnum>94</pnum>
      <trial>
         <trial_num>145</trial_num>
         <item>screwdriver</item>
         <patient_response>"driver-screw"</patient_response>
      </trial>
   </patient>
   <patient>
      <pnum>175</pnum>
      <trial>
         <trial_num>4</trial_num>
         <item>glasses</item>
         <patient_response>"glasseses"</patient_response>
      </trial>
   </patient>
</patients>

The previous results were generated by the following query: SeventhQuery

8)
Results: Corina query 8, 2D map view

The previous results were generated by the following query: EighthQuery

This question was interpreted as "give me all patients who exhibited a type 2 error (semantic) in a named anatomical region but did not exhibit a type 3 or 6 (phonological or phonological/semantic blend) in the same region, for each display the names and codes for all such regions".

9)
Results: Corina query 9, 2D map view

The previous results were generated by the following query: NinthQuery

This question was interpreted as "give me all patients who exhibited a type 2 error (semantic) in a named anatomical region but did not exhibit a type 8 (phonological reduction) in the same region, for each display the names and codes for all such regions".

10)
Results: Corina query 10, 3d MindSeer view

The previous results were generated by the following query: TenthQuery

The following procedure was used in this query: Gather all patients, their surgical stimsites, and the trials associated with those stimsites IF they resulted in an error code. For these if the stimsite coordinates, transformed into MNI space lie within specified deviations (deviations specified for each axis anterior, superior, and right) of a specified target coordinate, and the stimsite has error producing trials, show this patient and those trial codes which apply. The specific query was for (-32,18,59) with allowed deviations of 5 in each dimension. NOTE: one thing that is perhaps not evident from this query is that a patient MAY have results that are in the corresponding area, but not have a transformation to MNI space. In this case no results are returned.

Attachments