MétaCan
Menu
Back to cohort
Record W4308179942 · doi:10.2196/39443

Learning the Treatment Process in Radiotherapy Using an Artificial Intelligence–Assisted Chatbot: Development Study

2022· article· en· W4308179942 on OpenAlexafffundvenue
Nathanael Rebelo, Leslie Sanders, Kay Li, James C. L. Chow

Bibliographic record

VenueJMIR Formative Research · 2022
Typearticle
Languageen
FieldComputer Science
TopicAI in Service Interactions
Canadian institutionsPrincess Margaret Cancer CentreUniversity of TorontoYork UniversityUniversity Health NetworkToronto Metropolitan University
FundersCanadian Institutes of Health Research
KeywordsChatbotProcess (computing)Computer scienceMedicineWorld Wide Web

Abstract

fetched live from OpenAlex

BACKGROUND: In knowledge transfer for educational purposes, most cancer hospital or center websites have existing information on cancer health. However, such information is usually a list of topics that are neither interactive nor customized to offer any personal touches to people facing dire health crisis and to attempt to understand the concerns of the users. Patients with cancer, their families, and the general public accessing the information are often in challenging, stressful situations, wanting to access accurate information as efficiently as possible. In addition, there is seldom any comprehensive information specifically on radiotherapy, despite the large number of older patients with cancer, to go through the treatment process. Therefore, having someone with professional knowledge who can listen to them and provide the medical information with good will and encouragement would help patients and families struggling with critical illness, particularly during the lingering pandemic. OBJECTIVE: This study created a novel virtual assistant, a chatbot that can explain the radiation treatment process to stakeholders comprehensively and accurately, in the absence of any similar software. This chatbot was created using the IBM Watson Assistant with artificial intelligence and machine learning features. The chatbot or bot was incorporated into a resource that can be easily accessed by the general public. METHODS: The radiation treatment process in a cancer hospital or center was described by the radiotherapy process: patient diagnosis, consultation, and prescription; patient positioning, immobilization, and simulation; 3D-imaging for treatment planning; target and organ contouring; radiation treatment planning; patient setup and plan verification; and treatment delivery. The bot was created using IBM Watson (IBM Corp) assistant. The natural language processing feature in the Watson platform allowed the bot to flow through a given conversation structure and recognize how the user responds based on recognition of similar given examples, referred to as intents during development. Therefore, the bot can be trained using the responses received, by recognizing similar responses from the user and analyzing using Watson natural language processing. RESULTS: The bot is hosted on a website by the Watson application programming interface. It is capable of guiding the user through the conversation structure and can respond to simple questions and provide resources for requests for information that was not directly programmed into the bot. The bot was tested by potential users, and the overall averages of the identified metrics are excellent. The bot can also acquire users' feedback for further improvements in the routine update. CONCLUSIONS: An artificial intelligence-assisted chatbot was created for knowledge transfer regarding radiation treatment process to the patients with cancer, their families, and the general public. The bot that is supported by machine learning was tested, and it was found that the bot can provide information about radiotherapy effectively.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.006
metaresearch head score (Gemma)0.025
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Bench or experimental · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.006
Threshold uncertainty score0.032

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0060.025
Meta-epidemiology (narrow)0.0010.001
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0010.000
Science and technology studies0.0010.001
Scholarly communication0.0010.002
Open science0.0020.001
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0040.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.216
GPT teacher head0.491
Teacher spread0.274 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designBench or experimental
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations41
Published2022
Admission routes3
Has abstractyes

Explore more

Same venueJMIR Formative ResearchSame topicAI in Service InteractionsFrench-language works237,207