Can Machine Learning Models Detect and Predict Lymph Node Involvement in Prostate Cancer? A Comprehensive Systematic Review.
Eliodoro FaiellaFederica VaccarinoRaffaele RagoneGiulia D'AmoneVincenzo CirimeleClaudia Lucia PiccoloDaniele VertulliRosario Francesco GrassoBruno Beomonte ZobelDomiziana SantucciPublished in: Journal of clinical medicine (2023)
(1) Background: Recently, Artificial Intelligence (AI)-based models have been investigated for lymph node involvement (LNI) detection and prediction in Prostate cancer (PCa) patients, in order to reduce surgical risks and improve patient outcomes. This review aims to gather and analyze the few studies available in the literature to examine their initial findings. (2) Methods: Two reviewers conducted independently a search of MEDLINE databases, identifying articles exploring AI's role in PCa LNI. Sixteen studies were selected, and their methodological quality was appraised using the Radiomics Quality Score. (3) Results: AI models in Magnetic Resonance Imaging (MRI)-based studies exhibited comparable LNI prediction accuracy to standard nomograms. Computed Tomography (CT)-based and Positron Emission Tomography (PET)-CT models demonstrated high diagnostic and prognostic results. (4) Conclusions: AI models showed promising results in LN metastasis prediction and detection in PCa patients. Limitations of the reviewed studies encompass retrospective design, non-standardization, manual segmentation, and limited studies and participants. Further research is crucial to enhance AI tools' effectiveness in this area.
Keyphrases
- artificial intelligence
- positron emission tomography
- computed tomography
- machine learning
- prostate cancer
- systematic review
- lymph node
- magnetic resonance imaging
- pet ct
- big data
- deep learning
- contrast enhanced
- end stage renal disease
- ejection fraction
- case control
- newly diagnosed
- radical prostatectomy
- prognostic factors
- chronic kidney disease
- squamous cell carcinoma
- neoadjuvant chemotherapy
- meta analyses
- convolutional neural network
- dual energy
- label free
- mass spectrometry
- sentinel lymph node