Towards a more objective schizophrenia diagnosis



That might be helpful. They took an mri of my brain when I was first hospitalized. The scan looked normal I heard as in no obvious defects or growths. However while they were taking the mri I heard them saying that it was strange and how is it possible that I had so little activity going on in my brain. I felt a little offended at the time but my guess is that you’re brain stops working when you’re hearing voices and seeing things. When I asked them about seeing the mri later within the next day after they were surprised that i even remembered having one cuz I guess the stuff they had me take during the mri was supposed to Knock me out so bad that I should not have even remembered getting one


Hearing internal voices were classified wrongly as a diagnostic symptom of schizophrenia !
Whereas,the Dr can not diagnosing the voices by his ears or by the lab methods.
the Pronounced Phonetic Material in itself is a single character of the sz cause her/his self ,while the hearing of the pronounced voices is the personal auditory reaction"perception"

Both of pronounce process and hearing are running within the electrochemical pulse and expended in the place/time of the conscious mind,whereas the external audience are unable to monitor these processes at all,therefor this is NO medical diagnosis for the pronounced voice or heard, it means that ; you can not classifying both of voices/hearing as a diagnostic symptoms !

Simply,where from you know the person with sz hearing a voices or not ?
the person tells that- not the Dr or any medical diagnosis methods


It seems there are a lot of methods for getting around 80% accuracy - maybe they could combine 2 or 3 and get 95%+ and then that would be something.

Ex. MRI, EEG, blood biomarkers/genetic tests.



About the video above…

fMRI data converted to musical sound. Brain images are preprocessed into 20 distributed ensembles, “Independent Components,” and each is assigned a tone on a pentatonic scale. The loudness of each note corresponds to the intensity of activity in the corresponding regions of the brain. In this video, you can listen to soundtracks of healthy subjects alternating with schizophrenia patients. At the end, there are a few comments about statistical differences that may explain some of the audible variation.


I think the main problem with this approach is that the researchers presumably based the distinction between sza and bipolar vs sz on the diagnostic manuals and not on the magnitude of the observed differences with the brain imaging techniques. Since the DSM and ICD have been called into question as evidence-based taxonomies, this should be avoided.


That is a good point. If you have anemia, it is not diagnosed according to whether you feel crappy or not, it is by a specific blood measure.

Edit: and you can feel fine and have anemia. Source: has happened to me.


Does 80% accuracy mean there would be 10% false negatives and 10% false positives ?


No, it means 80% hits and 20% misses. I.e. (true positives + true negatives) / (true positives + true negatives + false positives + false negatives). It doesn’t say anything about the false positives vs false negatives ratio.


“80% accuracy in experimental conditions.”

Can anyone define ‘experimental conditions’?


I haven’t read the paper, just the article, but I’m assuming they just meant to point out that the experimental nature of the study could limit its generalizability. It’s a standard caveat that’s mentioned in almost all papers on experimental studies.


Thanks @Treebeard

Wasn’t sure what that meant.


No, sorry. They probably meant that the 80% accuracy was in the experimental group and not the control group. I can check when I get home.


I’ve always considered schizophrenia to be a syndrome, a group of symptoms that are chronic and stable over time.

Maybe these scans can conclude an objective diagnosis. I don’t know.


It would be objective as in the criterion is objective (patterns in EEGs). But objective doesn’t always mean better, and it doesn’t mean schizophrenia as a category is objectively valid.


First electroencephalogram (EEG) signals from 20 schizophrenic and 20 control subjects were acquired and then five bands’ power (standard bands) were elicited from each EEG channel. Discriminative bands were extracted using genetic algorithm (GA), particle swarm optimization (PSO) and ant colony optimization (ACO) and then were fed to Fisher linear discriminant analysis (FLDA) to classify the two groups. Experimental results demonstrate 83.74%, 81.41% and 81.06% classification accuracy for PSO, ACO and GA feature selectors, respectively

So basically, one group of people with a diagnosis of schizophrenia and one group of controls. It doesn’t mention what the characteristics of the control group were except that they were age-matched, so I’m assuming they were healthy controls and not actually people with schizoaffective or delusional disorder. This also means, @firemonkey, that the 80% accuracy is 80% true positives and 20% false negatives, since the entire experimental group had a previous diagnosis of schizophrenia.


I have simple question to you;
What is degree of accuracy for blood biomarkers/genetic tests while the person with sz (specimen) asleep ?


There are published studies on blood and/or genetic biomarkers in the 75-80% accuracy range. Presumably all the blood samples are taken while awake. However, your genome doesn’t change while you’re asleep.


You say: Your genome doesn’t change while you’re asleep.
This means that,if the Sz is a genetic disease,it will exists and work along 24 hours (waking/sleep)

You can have a new discovery in the Sz research,if you make a test for blood/genetic test while the person asleep ,during the time of deep sleep there is no Sz case/cause,effect ( interactions /reactions) or any type of change in the chemical balance state,cognitive state …etc ,as sz does not exists !!
this will proof to you that : sz is not biological/genetic/inherited in nature !
Just do that !