Adding classes to radont.owl

[Bayesianworld]

Created a new branch of radont.owl that contains a number of new classes.

• ionization chamber
• monitor unit
• frame of reference
• control point
• subclasses of external beam device [linear accelerator {tomotherapy, robotic, C-arm), cyclotron, synchrocyclotron]
• collimator

[Bayesianworld]

Some more additions:

• added several collimator subclasses
• added radiation_delivery_specification
• added detail to radiation_therapy_plan_specification
• Question: is a prescription an objective specification or a plan specification?

[Bayesianworld]

Added the annotation "rdfs:isDefinedBy" and my name to those classes that I defined. Seems to be the OBO thing to do.

[Bayesianworld]

Course of radiation: I was looking into whether we can/should. make such a class. According to OROO, the definition is "Grouping collecting all radiation therapy treatments constituting an episode of care. This means all target volumes treated from first delivery of radiation in the course until physician believes they have treated all body sites in need of radiation therapy. So if there is a split course for liver with 1 month between, these are all part of the same course. If treatment is started for one body site and then another is subsequently discovered and treated, the course includes both. For each instance the DateTime and Patient Age at the time of delivery of the first delivery of therapeutic radiation in the course is recorded."

In RTO, we could include it under "occurrent => planned process". However that class is a concretization of a plan specification. OROO's definition says that it can actually evolve during the process, i.e. dealing with complications, emergence of site of disease, etc. So there really is only the "plan of a plan specification". In our earlier meetings, we discussed how a course actually is a process that includes some logic, e.g. do x if y, but do z if not y.

Looks like we might best be served if we do not define it but give examples of how inference can be. used?

[Bayesianworld]

Continued adding some classes in the occurrent branch. Added OGMS bodily process and pathological bodily process. Used this as parent to radiation induced pathology => radiation complication => early and late effects.

[Bayesianworld]

Again, looking at OROO, I thought of staging as a class. Here are the attributes of the OROO Staging Class: Staging System ICD Version ICD Code ICD Code Qualifier Relevant Surgical Status Staging-T Category Staging-T Staging-N Staging-M Gleason Grade Group PSA Value Gleason Score Number Of Cores Biopsy Pathology Imaging for Biopsy Prior Active Surveillance Disease Extent Assessment Method Initial Prostate Volume Family History of Cancer Primary Diagnosis And Staging When Treating Secondary

As you can see, there is a lot going on here. There was also a recent article in IJROBP (Kim et al, Vol 112, Issue 1, 2022) about how gliomas are being re-staged based on genetic/biomarkers. Highlights difficulty in defining much about stage other than that it is an ordinal system related to some endpoint, usually survival. That might be all we want to include. On the other hand, we may also put in an IAO concept of stage and leave it that without going into how it comes about. That would allow correspondence with clinical EHS systems.

[Bayesianworld]

Added IAO classes for cartesian coordinate system and 1,2,3 dimensional subclasses.