Back

Performance of Handcrafted Radiomics versus Deep Learning for Prognosticating Head and Neck Squamous Cell Carcinoma: A Systematic Review with Critical Appraisal of Quantitative Imaging Studies

Gouthamchand, V.; AF Fonseca, L.; JP Hoebers, F.; Fijten, R.; Dekker, A.; Wee, L. Y.; Thomas T, H. M.

2024-10-22 health informatics
10.1101/2024.10.22.24315007 medRxiv
Show abstract

Head and neck squamous cell carcinoma (HNSCC) presents a complex clinical challenge due to its heterogeneous nature and diverse treatment responses. This systematic review critically appraises the performance of handcrafted radiomics (HC) and deep learning (DL) models in prognosticating outcomes in HNSCC patients treated with (chemo)-radiotherapy. A comprehensive literature search was conducted up to May 2023, identifying 23 eligible studies that met the inclusion criteria of methodological rigor and long-term outcome reporting. The review highlights the methodological variability and performance metrics of HC and DL models in predicting overall survival (OS), loco-regional recurrence (LRR) and distant metastasis (DM). While DL models demonstrated slightly superior performance metrics compared to HC models, the highest methodological quality was observed predominantly in studies using HC radiomics. The findings underscore the necessity for methodological improvements, including pre-registration of protocols and assessment of clinical utility, to enhance the reliability and applicability of radiomic-based prognostic models in clinical practice.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.