Time series radiomics for the prediction of prostate cancer progression in patients on active surveillance

Abstract

Serial MRI is an essential assessment tool in prostate cancer (PCa) patients enrolled on active surveillance (AS). However, it has only moderate sensitivity for predicting histopathological tumour progression at follow-up, which is in part due to the subjective nature of its clinical reporting and variation among centres and readers. In this study, we used a long short-term memory (LSTM) recurrent neural network (RNN) to develop a time series radiomics (TSR) predictive model that analysed longitudinal changes in tumour-derived radiomic features across 297 scans from 76 AS patients, 28 with histopathological PCa progression and 48 with stable disease. Using leave-one-out cross-validation (LOOCV), we found that an LSTM-based model combining TSR and serial PSA density (AUC 0.86 [95% CI: 0.78–0.94]) significantly outperformed a model combining conventional delta-radiomics and delta-PSA density (0.75 [0.64–0.87]; p = 0.048) and achieved comparable performance to expert-performed serial MRI analysis using the Prostate Cancer Radiologic Estimation of Change in Sequential Evaluation (PRECISE) scoring system (0.84 [0.76–0.93]; p = 0.710). The proposed TSR framework, therefore, offers a feasible quantitative tool for standardising serial MRI assessment in PCa AS. It also presents a novel methodological approach to serial image analysis that can be used to support clinical decision-making in multiple scenarios, from continuous disease monitoring to treatment response evaluation. Key Points •LSTM RNN can be used to predict the outcome of PCa AS using time series changes in tumour-derived radiomic features and PSA density. •Using all available TSR features and serial PSA density yields a significantly better predictive performance compared to using just two time points within the delta-radiomics framework. •The concept of TSR can be applied to other clinical scenarios involving serial imaging, setting out a new field in AI-driven radiology research.

Prediction of Breast Cancer Diseases From Genes Using Pso and Cso Tuned Long Short-term Memory

“…Classifiers with five different feature selection are used to identify relevant features from microarray cancer datasets [6]. To improve accuracy in prediction of prostate cancer, the study employs LSTM recurrent neural network (RNN) and creates Time Series Radiomics (TSR) predictive model [7]. Ensemble classifier approach fine-tunes the critical parameters in each layers of Network [8].Hybrid reduction technique is used to optimizes the process for breast cancer diagnosis [9].Fuzzy C-Means Clustering is applied to optimizes the parameters in RNN through CSO for lung cancer detection [10].A non-invasive breast cancer classification system for Metastatic Breast Cancer (MBC) diagnosis was proposed through ML models using blood profile data of MBC patients for survival prediction [11].Deep Optimal Neurocomputing Technique through Multivariate Analysis is used to predict the cancer which reduces the computation time [12].A innovative technique, termed as self-adaptive sea lion optimization algorithm is used to optimize the weights using cutting-edge meta-heuristic algorithm [13].Adaboost and ensemble machine learning technique predicts early lung cancer.…”

“…o_n = σ(z_o * [g_(n-1), x_n] + b_o) (6) Candidate cell state (D_t) data is kept in the cell state and represented by the candidate cell state (D_n). The hyperbolic tangent (tanh) triggering function calculates and reduces the range of values between -1 and 1.The candidate cell state's mathematical equation is shown in equation (7).…”

“…D~_n= tanh(Z_d * [g_(n-1), x_n] + b_d) (7) Current cell state (D_n) update, Combies the data from the candidate cell state (D_n), forget gate (f_n), and prior cell state (D_(n-1)), final cell state (D_n) is calculated. The cell state updates mathematical equation through equation (8).…”

Novel efficient reservoir computing methodologies for regular and irregular time series classification

“…Over the past few decades, an astronomical amount of algorithms have been developed to address this particular field. Thus far, Long Short-Term Memory (LSTM) networks can be seen as a milestone breakthrough, offering a robust solution to the challenges posed by modelling complex long-term dependencies in sequential data [4][5][6][7]. LSTM networks are a type of recurrent neural networks (RNN) that take advantage of memory cells and gates as a means to control the flow of information through the network so that the vanishing gradient problems can be largely mitigated.…”

Using a Recurrent Neural Network To Inform the Use of Prostate-specific Antigen (PSA) and PSA Density for Dynamic Monitoring of the Risk of Prostate Cancer Progression on Active Surveillance

“…BGV for PSAD NA and follow-up gland volumes for PSAD A were calculated from MRI scans according to Prostate Imaging-Reporting and Data System guidelines [7] using three-plane measurements by four consultant urogenital radiologists with 4–14 yr of prostate MRI reporting experience. A previously described [8] long short-term memory recurrent neural network with leave-one-out cross-validation was applied to the data to generate areas under the receiver operating characteristic curve (AUCs) for predicting PCa progression on AS. Progression was noted in 80/332 patients, defined as either histopathological (biopsy-confirmed International Society of Urological Pathology grade group upgrading) or clear radiological stage progression (PRECISE [9] score of 5).…”