Categories
Information Technology

Free Sample of Upcoming Book (Label Normalization)

Hi All!

Please see below a sample section of my upcoming book on Label Normalization (sample may be updated and book may be different from sample). I am slowly releasing all my knowledge and work on alternative psychiatry and psychiatric computing. Some of my work will be for sale but a lot of it is being given away for free. The book on Label Normalization will be a book for sale however this is a free sample.

Label Normalization is all about removing a lot of the unnecessary data that patients are associated with and in particular, how information technology can facilitate label removal/non-dependence (meaning you don’t have to deal with psychiatric labels anymore). I know that sounds strange but it’s true and so I hope you enjoy this sample and it satisfies your curiosity!

If you like my work, please consider visiting the Donate page. This helps me get my work out and helps in many other ways I appreciate it! Thanks!

SAMPLE BELOW

Is label non-dependence (label removal) even possible? if so, how?

Verifying whether label non-dependence is possible or not is actually a very simple task. It is actually common sense and can easily be established. A simple example is demonstrated with the depressed mood of depression and the depressed mood of bipolar disorder. Psychiatrists issue anti-depressants for the depressed mood found in both labels. The fruit of a psychiatrist’s work is to offer treatments that target such negative symptoms. But labels such as depression and bipolar disorder are not a necessity when it comes to anti-depressants as they are simply nouns that host a set of symptoms. Anti-depressants will work against the symptoms regardless of whether a patient uses a label or not. In other words, the drugs target the symptoms by manipulating serotonin in the brain and they do this without dependence on a label (obviously). This is because labels are in the domain of language but symptoms and the drugs that are used to treat them are in the domain of medical conditions and chemistry. Without symptoms, labels are useless. Therefore labels are not a requirement in order to manipulate the brains neurotransmitters and thus induce therapeutic outcomes.

Even though label-non dependence is a straight forward and common sense concept, it doesn’t seem to be applied in the real world which is evidenced by the lack of innovation in psychiatry in this regard and the amount of dependence on numerous labels used by both practitioners and patients. But using label non-dependence wisely in an organized and structure manner actually gives birth to a very powerful set of innovations, ideas and technologies. Whilst label non-dependence is a simple concept, to see its full potential and power in action, it is necessary to establish it in an organized way such as using it in combination with information technology. Thankfully, information technology can do very complicated things that we can’t just like a camel can take a traveller through a hot desert. Information technology handles very complicated things in a very simple way.

There are various established scientific classification/categorization models widely in use today that are very applicable to psychiatric disorders. However, the current psychiatric establishment does not utilise such systems in regard to diagnostic classification. These systems when applied to psychiatric disorders clearly demonstrate their ability to facilitate label removal and non-dependence amongst many other powerful benefits. In the world of computing, there are many of these modelling systems and some of them seem very similar, consistent and connected to each other. To begin describing the concept of alternative classification in psychiatry and how such a concept facilitates label removal, I will start with a database design system known as the “Entity-Relationship Model”. I am starting with this modelling system as it gives very good fundamental knowledge on natural classification regardless of the domain.

Linguistically, when we want to classify or name something we use words like “entity” or “class”. Both these words are heavily used in the computing world. The word entity is simply a noun which refers to anything with distinct and independent existence. The following are examples of entities:

  • Animal
  • Bank Account
  • Cloud
  • Computer
  • Human
  • Etc

The word entity by itself does not indicate any specific thing until someone specifies exactly what the entity is. For example, if someone says, “our universe Is full of entities”, they are simply saying our universe is full of “things” with their own distinct and independent existence. As there was no specification of what entities the speaker was referring to, there is general reference to whatever can be seen in the universe whether it be stars, the moon, plants, animals etc. On the other hand, if someone says, “the sun is a very bright entity”, in that case the entity has been specified as the sun. I mentioned this point of entities, generalization and specification because it is important to know for alternative classification of psychiatric disorders. From what I observe, psychiatry doesn’t use a natural classification system which is why I feel the DSM is a mess and contains a lot of repetition and redundancy. Natural classification however involves general and specific entities and this is easily observable in the natural world.

The use of the word entity is similar to the use of the word “thing”. We use the word thing to give a quick and easy identification, classification or name to something abstract, general, unknown and sometimes even specific. You can actually use it to refer to anything which is what makes it so useful. This talk of entities and relationships is nothing philosophical or an issue worthy of debate, it is proven and based on language, nature and what we see in the observable world. This is why, as mentioned below, the information technology industry has adopted something called the ER model which stands for the Entity Relationship model.

In software engineering, database engineers use something known as an Entity-Relationship model (ER model for short) to aid in database design. The ER model was developed by Peter Chen and published in his 1976 paper “The Entity Relationship Model – Toward a Unified View of Data”. Pictured below is an example of the way entities and their relationships are expressed through the ER model.

In the above image, there are two entities which are Employer and Employee. Right at the top of the entity in dark blue is its title (Employer or Employee). Below the title cell, the attributes for each entity are shown. The top attribute highlighted in light blue is known as a “primary key”. Primary keys are used to guarantee uniqueness as they are what make an entity unique, but such keys are beyond the scope of current discussion. The red line you see between the entities represent a relationship between them. In this case, an employee has a link to an employer because every employee belongs to a certain employer. In the image this is why we see the EmployerID attribute in both the Employer and Employee entities and the red line pointing from the EmployerID in the Employee entity to the EmployerID of the Employer entity (i.e Employee is foreign to Employer but they are associated because every employee is assigned an ID which belongs to and identifies an employer). Yikes! What a mouth full! I know, but it’s not important to know this much technical detail right now so let’s move on.

In the below series of images, I demonstrate another expression of the ER Model except the context relates to psychiatric disorders and label removal (non-dependence). You will therefore see label removal in action. The first image below is known as the pain base class abstraction. This abstraction is basically a simple generalization that describes the nature and anatomy of every psychiatric disorder. It is necessary to display this abstraction first as all the following images stem from it.

The pain base class abstraction above describes the anatomy of every psychiatric disorder at its fundamental level. This abstraction is based on a design language used in the field of software engineering known as UML or the Unified Modelling Language. More specifically, the above abstraction has been inspired by something known as a class diagram which is found within the UML standard. The UML specification contains many different diagrams with class diagrams being one of them. Class diagrams are used to design software systems that are constructed through a design paradigm known as the Object-Oriented paradigm. This paradigm (design standard) has contributed to the establishment of some of the biggest software and web systems that have changed the world.

The image below is a translation of the aforementioned pain abstraction into the ER model standard using psychiatric diagnostic data.

Label refers to psychiatric labels, disturbance is just another word for symptoms and the third entity in the middle is known as an “associative entity”. The LabelDisturbance associative entity connects labels and symptoms and allows major computational manipulation of diagnostic data which will be shortly demonstrated. An object-oriented design such as the pain abstraction does not always translate verbatim (directly) when it enters the entity relationship world. For example, the “communicate” behaviour at the bottom of the abstraction actually represents an abstraction of all possible symptoms for all possible psychiatric disorders. When translating this abstract behaviour to the ER model, the three entities Label, Disturbance and LabelDisturbance are naturally generated. However Label and LabelDisturbance were only generated for convenience to communicate with psychiatric diagnostic data. They are not a necessity in alternative classification.

The above ER model is the basis of a valid database design and so in the following images, I will demonstrate the use of SQL (Structured Query Language) to manipulate psychiatric diagnostic data and do things like remove labels whilst retaining symptoms and thus show how labels are redundant data. Please note, in our sample database, I have only inputted three labels and each label may not contain their full list of official symptoms. This is because my goal is to only demonstrate a concept.

The following SQL statement retrieves all three labels in the database with their symptoms in the adjacent column:

In the next query, we will query the database and request all labels with their symptoms again, but this time also show the associated abstract pain class that the symptoms (disturbances) derive from and are linked to. Recall that all symptoms (disturbances) were abstracted under the communicate behaviour of the pain abstraction. Therefore, the next query must show a link to the pain abstraction which possesses the communicate attribute.

The reference to the Pain abstraction shown in the previous result set is made possible because the abstraction is logically associated with symptoms (disturbances) as indicated in the ER model below:

If you didn’t understand where the pain abstraction is in the above image, it’s within the class and class property entities which you will understand through the images below:

The following query extracts the pain class from our database.

The “ClassProperty” entity in our database holds the remaining details of the pain abstraction class. To retrieve those remaining details, we need to send the following query to the database:

Now it’s time to get into the exciting stuff. How do we officially separate (decouple) psychiatric labels from symptoms and thus induce label non-dependence whilst still retaining a patient’s association to a medical condition? And how do we ensure this happens in an organized and safe way that is governed by an established science? We simply send the appropriate query to the database.

First recall result set 2 again pictured below:

In our new query, like in the query that produced result set 2, we will again request the database to return the symptoms associated with the three labels schizophrenia, bipolar disorder and OCD. But this time however, we will hold all the symptoms under the pain class abstraction and discard the labels the symptoms are associated with. We can do this without any loss to the patient because as mentioned earlier, disturbances (symptoms) naturally derive from the pain abstraction’s communicate behaviour. Please see the below query and result set:

As can be seen above, the labels schizophrenia, bipolar disorder and OCD have become disconnected to their symptoms (label-symptom decoupling) and have been generalized under a single category called “Pain”.

The main goal of a practitioner towards a patient in psychiatry is to offer effective treatment to hopefully assist them in achieving a normal and healthy life. For this to occur, practitioners target and treat the symptoms that ail their patients. The images above demonstrate the achievement of label non-dependence which is indicated by the absence of labels in the result set. From a treatment perspective, based on the result set, a practitioner can still target and treat the patient’s ailment as the symptoms have been retained despite the labels being removed. The retained symptoms are the values situated in the DisturbanceName column. 

The pain class naturally defines, at a more general level, what type of entity a disturbance is. This is how it is able to replace/discard the labels in the result set above. Symptoms in other words are a type of pain but at a more granular and specific level. The moment someone starts talking about symptoms, they are simply discussing the details of a type of pain. There is no loss of quality by applying this new classification system to psychiatric diagnostic data. Rather, on the contrary, it is the application of a more natural and powerful classification system already in high use elsewhere in the business world.

The following image highlights the link between the pain abstraction and the symptoms displayed in the above result set. I have highlighted the communicate behaviour property as this is where symptoms are held.

Our ER model is at the same level of our query and result set. Therefore, to provide a visual on how the result set was achieved, the below image is the ER model view of the relationship between the Pain class and symptoms.

The question may arise, “don’t labels allow us to differentiate between symptoms? isn’t that their whole purpose? doesn’t that prove that we need them?”. It’s a good question, I mean that’s why labels are there, they are names given and attached to a set of symptoms that the psychiatric establishment consider a distinct and independent medical condition. My response is that my goal in writing this section was to prove a point and to demonstrate that labels are not a necessity. Labels not being a necessity opens up a whole new world of opportunity in regard to treatment, the way patients and practitioners see things, and progress in the industry of psychiatry.

Consider the two tables below. One shows OCD as psychiatry views it and the other shows essentially the same thing but without the label. This clearly demonstrates that labels are not a necessity and patients have the liberty to lose their attachment to them.

Regardless, the object-oriented paradigm accommodates the need for distinction very powerfully as it is a natural classification system. Concepts such as sub classes, inheritance and polymorphism may be some further reading for you if you would like to know more. I myself acknowledged in my own psychiatric classification system that there was indeed some need for distinction and to go into more details. Again, as mentioned above, my intention behind this discussion was to provide a brief introduction to the concept of label non-dependence. To demonstrate label non-dependence simply, via the object-oriented paradigm and information technology, I didn’t need to go further than the pain class abstraction as it is the fundamental base description of every psychiatric disorder. Using more than the pain abstraction to explain label non-dependence would have made it harder to understand and would be considered overkill. The fact that it was sufficient for me to use the pain abstraction to get the point across also demonstrates the power of the object-oriented paradigm as a natural alternative to psychiatry’s DSM as I was able to use a simple high abstract view of disease together with symptoms that are more granular and this is because of the natural connection between the two extremes.

Distinction and separating conditions are another topic of discussion. Differences are obvious though, for example, the symptoms of social anxiety disorder are clearly different to what is understood to be obsessive compulsive disorder. But that doesn’t mean a generalized label of “Pain” isn’t a valid title for both. It is naturally a valid title which is why a practitioner could work with such information. The only distinction is at the symptom layer.