bioRxiv · 20 Nov 2026
Biologically Guided Deep Learning for Personalized PSA Forecasting in Intermittent Prostate Cancer Therapy
Shreya Yashodharh-index 0, Kevin Leder
Citations by database
Each database indexes a different set of journals, preprints and books, so they are shown side by side rather than blended.
- OpenCitations0
- Crossref0
- Europe PMC0
Citations per year
The current year is still counting.
- 2026 so far0
Abstract
Intermittent Androgen Deprivation Therapy (IADT) offers a promising management strategy for prostate cancer by alternating treatment periods with the aim of delaying castration resistance and mitigating side effects. An evolutionary rationale for treatment interruptions is to preserve treatment-sensitive cells that may competitively suppress resistant populations. However, designing optimal patient-specific treatment schedules remains challenging due to the complex dynamics of tumor evolution and the computational intensity of traditional mechanistic simulation models. Standard mathematical models have prespecified structures that may inadequately capture patient-specific tumor evolution, limiting Prostate Specific Antigen (PSA) forecasting accuracy and adaptation to new patient data. These limitations may reduce the effectiveness of treatment policies learned using these models. To address them, we present a domain-adapted machine learning framework that integrates biological constraints into the deep learning architecture and embeds historical biomarkers to act as a robust, computationally efficient neural surrogate for mechanistic simulations. This framework enhances generalization across sparse clinical profiles by leveraging deep neural representations for patient-specific and cross-patient parameter learning, and potentially enables the identification of personalized dosing schedules tailored to individual tumor trajectories. Preliminary evaluations confirm that it reliably captures treatment-associated PSA dynamics while overcoming the computational bottlenecks of numerical solvers. Ultimately, this work provides a scalable foundation for optimizing adaptive therapy protocols to extend time to disease progression. We present the architectural formulation, forecasting evaluation against mechanistic baselines, and training and optimization procedures of the deep neural networks.
What the data shows
- No database records a citation yet, as is usual for a paper this new.
- Shreya Yashodhar has the highest h-index among the authors (0).
Graph
This paper in the middle, its authors and their most-cited papers below, the papers it cites to the left and the papers citing it to the right. The timeline lays every paper out by year.
Select a node for its details. Drag to pan, scroll to zoom, drag a node to move it.
Start with
Cited by 0 papers
Who cites this paper, from open citation data, with Crossref's record of each; the 100 newest when there are more.
References (0)
What it cites, from open citation data, with Crossref's record of each.