System radiobiology modelling of radiation induced lung disease
Bibliographic record
Abstract
Radiation induced lung disease (RILD) is a side effect of radiotherapy for treating thoracic cancers, limiting radiation dose to tumours and in turn the chance of treatment success.A current scheme for predicting and managing RILD risk is based on a population-based normal tissue complication probability (NTCP) model assuming the same response to given radiation dose in lung.However, recent research suggests that dose response can be modified by biological and clinical factors pertinent to pathogenesis of RILD.In this work, we explore systems radiobiology approaches to model RILD as a result of interactions between these factors.Clinical, dosimetric, and biological data on lung cancer patients were analyzed to identify markers associated with high RILD risk.Then, we applied machine learning methods to combine such markers into models that calculate patient-specific RILD risk.We investigated two RILD endpoints: radiation fibrosis (RF) and radiation pneumonitis (RP).RF is formation of scar tissues in lung and can be quantitatively measured from computed tomography (CT) images.We extended a classical NTCP model to explicitly model time-dependent dose response of RF risk.Our modelling results have shown significant change in dose-RF correlation after 3 months post-treatment as well as higher RF risk when tumour was in lower lung.We extended the dose modelling to intra-treatment CT images.However, we did not find association between early CT changes and biological states or clinical outcomes.Subsequent investigations on radiation pneumonitis (RP) also suggest that dose response is modified by factors not related to lung dose distribution, such as dose to heart or production of proteins i guidance that he carried out with his intelligence and dedication.Meeting him the first time in May 2009 -the day he delivered an interview presentation about his vision on systems radiobiology -was a turning point of my academic career.That was the time he brought me into the world of machine learning which was obscure to most of us in this field.I see myself so lucky to having conducted this truly interesting research under his direction.Second, I would also like to thank my co-supervisor Jan Seuntjens for supporting the biomarker protocol as well as his leadership in the department which brought the CREATE program.Many thanks to Norma Ybarra for all the hard work in the wet lab and educating me with her endless knowledge in biology.Asha Jeyaseelan, my "the other lung", worked diligently for recruiting patients and setting collaboration with CHUM, which she did exceptionally well with her charm.Thank you Seema Ambereen for continuing Asha's work to keep this project progressing.So many thanks to Neil Kopek and Nathalie Japkowicz for their expert advices and commitment in the thesis committee.The CT imaging projects would have been impossible
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 imitationNot 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.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.006 | 0.001 |
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.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".