Dependency of a validated radiomics signature on tumor volume and potential corrections
Bibliographic record
Abstract
640 Purpose: Radiomics is foreseen as an essential prognostic tool for cancer risk assessment. The radiomics signature described in Aerts et al. (Nat. Commun. 2014) contains four features (1st order intensity, geometrical shape and higher order texture in image and in the wavelet domain): energy, compactness, grey-level non-uniformity (GLNU), and GLNU_HLH (GLNU in high-low-high subband). This signature was derived from pre-treatment CT scans of 422 lung cancer patients and subsequently tested in independent lung and head and neck (H&N) cancer cohorts (concordance-index of 0.65 in lung and 0.69 in both H&N cohorts). The authors claimed they had “shown for the first time the translational capability of radiomics in two cancer types” and that “radiomics quantifies a general prognostic cancer phenotype that likely can broadly be applied to other cancer types”. We investigated the dependency of the signature on tumour volume and suggest modifications to reduce it. Material and Methods: We extracted the features of the radiomics signature from both PET and CT components of PET/CT images of a multi-centric (4 different hospitals in Quebec) cohort of 300 H&N cancer patients. Absolute Spearman rank correlation (rs) was calculated between each of the four features and tumour volume. In order to reduce the observed dependency of the radiomic signature with tumour volume, we suggest alternative calculations for the shape and higher order features (see supplemental material). Results: Rs in PET (respectively CT) were 0.73 (resp. 0.71), 0.98 (resp. 0.94), 0.99 (resp. 0.98) and 0.89 (resp. 0.95) for energy, compactness, GLNU and GLNU_HLH respectively, showing great dependency of the radiomics signature on tumour volume. Using the revised calculations led to significantly lower rs values in PET (resp. CT) of 0.57 (resp. 0.50), 0.31 (resp. 0.15) and 0.11 (resp. 0.18) for compactness, GLNU and GLNU_HLH. Conclusions: Although the prognostic value of the radiomics signature was demonstrated in two cancer types in Aerts et al., we suggest that this translational capability of radiomics may have been primarily due to the fact that tumour volume is a strong prognostic factor in both cancer types. Indeed, according to the supplemental material of Aerts et al., concordance-index of volume alone was very close to that of the radiomics signature with 0.63 in lung and 0.68 and 0.65 in the two H&N cohorts. Further work is warranted to verify the translational capability of radiomics to different cancer types. Our first step will consist in evaluating the prognostic and translational value of the signature with revised calculations in both pathologies.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 teacher head, 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".